Domain Specific Software Architecture for Design Center Automation

Domain Sp eciﬁc Soft w are Arc hitecture for Desi g n Cen ter Auto mation Ansh uman Sinha, Haritha Na ndela, Vija y a Balakrishna Jan 2000 Abstract Domain sp e ciﬁc softwar e ar chite ctur e aims at softwar e r euse thr ough c onstru c- tion of domain ar chite ctur e r efer enc e mo del. The c onstru cte d r efer enc e mo del pr esents a set of individual c omp onents and their inter action p oints. When starting on a new lar ge s oftwar e pr oj e ct, the design en gine er starts with pr e- c onstructe d mo del, whic h c an b e e asily br owse d and picks u p opp ortunities of use in the new s olut ion design. This r ep ort discusses applic a t ion of domain r ef- er enc e design metho ds by deriving domain sp e ciﬁc re fer enc e ar chite ctu re for a pr o duct or deri n g system in a design c en t er. The pr o duct in this c ase is insto ck and sp e cial or der blinds fr om diﬀer ent manufactur ers in a lar ge supply stor e. The development of matur e domain sp e ciﬁc r efer enc e softwar e ar chite ctur e for this do main is not the ob je ctive of this r ep ort. However, this r ep ort would like to c aptur e the metho d use d in one such pr o c ess and t hat is the primary c onc ern of this r ep ort. This r ep ort lists su bje ctive detail s of such a pr o c ess applie d to t he do- main of or dering cust om and insto ck blinds fr om a lar ge home c onstruct ion and go o ds supply st or e. This re p ort also de scrib es the detaile d pr o c ess of derivation of know le dge mo dels, u niﬁe d know le dge mo dels and t he r efer enc e ar chite ctur e for this domain. However, this domain mo del is only p artial ly c omplete which may not b e use d for any r e al applic atio ns. This r ep ort is a r esult of a c ourse pr oje c t undertaken while studying this met ho dolo gy. Chapter 1 Problem Domai n 1.1 Domain Description The design center is a one-stop shop for customers to c ho o se, buy and install window-treatmen ts, wallpap e r, carp et, and tiles. Typical customers of the De- sign Center are homeowners, business communit y , a partment co mplex es etc. The foc us o f this pro ject is on the diﬀerent activities of the Design Center at Home Depo t. The g oal is to identif y these activities o f the Design Center, such as pro duct selection, Order Maintenance and T rac k ing , Billing , Sch eduling of app ointmen ts for measurement, installation and delivery , etc. Cho osing diﬀerent pr o ducts like wallpap e r, blinds, etc. can b e co nfusing, The custo mer is pr ovided with m ultiple c hoice s of pro ducts to choo s e from. This selection pro cess emphasizes a dhering to issues like: customer’s budget, time constraints, sp eciﬁcatio ns a nd expectatio ns . The selectio n pro ce ss de- pends on many diﬀeren t cr iteria like the dimensions, type of pro duct, part of the house i.e. kitch en, family r o om etc., c o lor, style the ro om p ortr ays, location of the ro om ea st, west etc., humidit y , color s surrounding the pro duct. There are diﬀerent products be ing manufactured by man y manufacturers. The categoriz a - tion of these pr o ducts into meaningful groups ca n hasten the sele c tio n pro cess . Employ ees of the design-center issue m ultiple quotes for the same categ ory of pro ducts in or der to provide the customer with diﬀer e nt optio ns to choo se from. Automation ha stens this selection pro cess and decision making b y the cus- tomer to by ha stening the decision. O n the other hand, automa tion can help the desig n cen ter employees in issuing c o rrect quo tes for diﬀerent s elections of the cus to mer. The desig n center sto cks many v a rieties o f pro ducts fr o m diﬀerent manufac- turers. Usually ther e is tw o-weeks w orth of materials in sto ck. As and when Home Dep ot introduces a new pro duct into the market they do this on based on 2 districts. Here they test the pr o duct sales a nd decide on an appropr ia te amount of inv entory to ma intain which is known as Suggested Order Q uantities (SOQ). Pro duct s a les are forecasted a ppr opriately to match the sales during the weeks to follow. As the pr o ducts are sold, the inv ent ory is up dated a s well. Whenever the in vent ory dips below the forecasted amount, the pro duct is reorder ed. Home Depo t also needs to tr ack the s ales in p erio dically so that ov e r -sto cking do es not o ccur. Another t y pe of rep ort that inﬂuences the in ven tor y is the velocity rep orts - which gives an idea abo ut the pr o ducts sold well. The design c e n- ter employ ee needs to ha ve a ll this informa tio n on ha nd, which improv es their eﬃcacy . One of the activities that ta kes place at Home Depo t or any other retail sto re is the re turn of go o ds. Custom order ed materia ls that are returned will not b e refunded unless there is a r eally v alid reason. Other pro ducts a re refunded if returned within the thirty days. When an item is returned to the stor e it is marked down to zer o a nd the pro duct is treated as a loss. Also the materials on the ﬂo or which are damaged, are als o marked do wn i.e. treated as a loss, and all this informa tion is r ecorded. After the items are mar ked down, they are sent back to the vendor (R TV). Automating this inv entory manag ement will ensure that the r equired amount of in vent ory for each pro duct is maintained. Currently the marked do wns that ar e trac ked using paper . This also helps in maintaining accura te vendor information, pro ducts oﬀered by that pa rticular vendor, maintain the de livery times and quality of serv ice provided by the ven- dor. Automation helps incr ease the eﬃciency and the spe e d of each tr ansaction. When a customer places an or de r with Home Dep ot, if the pro duct is in sto ck, the customer buys the pro duct. On the other hand if this is a custom order , the or de r deta ils are entered into the system and these o rders are automa tically faxed to the v endor s (supplier). The vendors pro cess these orders and supply the require d mater ial. Automation will help eliminate the err ors in the purchase order by ensur ing that the employ e e enters all the information nee de d to place the order. It a lso prompts the employ ee for a ny required accesso ries or additional ﬁttings that enha nce the selec ted pro duct. Depending on the inv entory , the order may need to b e placed to the manufacturer of the pro duct. Any p ossible changes r equested by the customer a re incor po rated into the o rder. One of the activities that take place at Home Dep ot is the Customer Ser v ice. Most of the pr o ducts sold at the Design center hav e some so rt of warrant y – either a tw o y ea rs or a lifetime. When cus tomer rep or ts a pro ble m, emplo yees are pre sent to r epair and reinstall the pr o duct. Home Depot also replaces the pro duct with no additional c harge. In many circumstances a third par t y co n tractor is sent from the design center to mea sure the dimensions for the blinds, wallpaper, carp et etc. This in volves int e r action with the customer to eﬀectiv ely determine the c lie n t’s needs. After the pro ducts are selected they will be orde r ed and H o me Dep ot will let the contractor know when the supplies a rrive. App ointmen ts need to b e made for delivery and insta lla tion. Automating this task ensures that the sc heduling 3 pro cess is more o rganized a nd eﬃcient. Ma jor issues r equests identiﬁed: 1 1. One of the is sues is the av a ilability of diﬀerent types of pro ducts and their classiﬁcation. 2. Custo m o rders or Inv entory repla cement o r ders normally take tw o weeks to deliver. But so metimes there is delay and this cause s the missing of the installation date. 3. One of the re q uests was that if a ll the co mputers in all Home Depo ts were linked there would b e easy access to information. 4. Also if there co uld an easy w ay of acces sing the vendor infor mation with pro duct av ailability . 1.1.1 Curren t Lev el of Automation Currently Home Depot has diﬀerent systems for diﬀere nt pro c e sses. They all use a mainframe system, whic h handles the or der placement, billing e tc. They hav e a mobile ca rt, using which or ders fo r in ven tory replace ment a re placed. A pro cess like updating the in ven tor y with the marked down pro ducts i.e. dam- aged pro ducts, is still tracked on paper and ha nded to the co mputer ro om for data en try . The design center e mployees need to know all the business rules to complete a s ales transaction. And to complica te the situa tion, most of the systems that exist are not user friendly . The selection pro cess is not a uto mated at present and this forces the cus- tomer to lo ok through all o f the diﬀerent catalo gs in trying to decide which one to cho ose. If this pro cess is automated, then the design center employ ee ca n assist the customer in a more eﬃcient wa y . When a customer s e lects a pro duct, it takes v a luable time in o r der to ge ne r ate a quo te. With the prop osed a utoma- tion, the quo te generatio n pro ces s ca n be hastened. Automa tion can make or der placement fas ter. As the order size incr e ases this feature b ecomes imp or tant . Currently the pro cess dealing with the return of go o ds is not automa ted. An exa mple 2 given b y the exper t was that the or der management pro cess is semi-automated i.e. quotes still need to entered one b y one for ea c h window instead of doing a bunch o f them together. So curre ntly there is automation in generating a quote but not to the extent req uired. 1 These issues ha v e been iden tiﬁed in an i nformal discussion wi th the exp ert. 2 This inform ation w as gained when w e had an inform al discussion about the domain description with the exp ert. 4 Business Pr o c ess Automated Semi-Automated Non Automated Selection * Order Mana gement * Inv entory Manag emen t * Billing * T able 1.1 : This table shows cla ssiﬁcation o f diﬀerent pro c e sses and their present level o f automation. 1.2 Domain Men tal Mo del and Viewp oin ts The Domain Men tal Mo del gives a g raphical description and the ﬂow of even ts that o ccur at the Home Dep ot. The ﬁg illustrates the diﬀerent pro ces ses in the men tal mo del. The following details describ ed b elow hig hlights the po ten tia l activities thr o ugh the mental mo del. 1. The customer ar rives at Design C e nter with requiremen ts for windo w treatments a nd/or wallpap er. 2. If the pro duct is wallpap er, the customer s hav e a n idea of their needs. If not they consult the employ e e for advice and make a selectio n after lo oking at diﬀerent catalo gs. 3. F or window treatments the customer wha t t yp e o f blinds do es he need the desig n cen ter employ ee assists the customer in dec iding on the type depe nding o n the customer r equirements. 4. F o r window treatments after the customer decides the t yp e an app oint- men t is made for measuring the dimensions a nd num b er o f window treat- men ts desired. 5. With the dimensions measure d, the customer then consults with a de s ign center emplo yee to decide on diﬀerent bra nd choices a v aila ble for them. The employee gener ates price quotes for diﬀerent choices the customer ha s made. The custo mer ﬁnally selects the pr o duct after considering diﬀerent asp ects. The vendor is co n tacted to ﬁnd if the pro duct is still in the market for the pro duct. 6. A t this moment the customer can still withdraw from the transaction if he do e s not like the choices oﬀer e d to him 7. If the particula r brand is in s to ck, the customer purc ha ses the pro duct. The inven tory is then up dated. 8. If the pro duct is not in s tock an order is placed with the v endor and the customer collects after its arr iv a l. Usua lly the vendors ha ve a t wo-week delivery date after the pro duct is r equested 5 Figure 1 .1: Domain Mental Mo del 6 9. If the pro duct is in sto ck then an o rder is placed, the cus tomer is billed for this purchase and after the paymen t is ma de the inv entory is up dated to reﬂect the change. 10. If the pro duct is not in sto ck then an or der is placed with the vendor, the customer is billed for the purchase a nd when the or der ar rives at the warehouse the cus to mer is notiﬁed. 11. Ther e are a couple of choices av ailable the customer can request the pro duct to b e delivered and installed or the customer can pick the pro duct and install it. 12. In any case if needed an app ointmen t is scheduled so that a p ers on can deliver the pur chased go o ds and install the same. 13. The terminating even t for the mental mo del is when the product meets the cus to mer sp eciﬁcatio ns. The diﬀerent viewp oints we need to consider a r e: the customer’s , the design center employ ee’s, the installer’s and the vendor’s. F rom the customer ’s view- po in t, the ex per ience of buying the pro duct s hould be as ea sy and painless as po ssible. The Design center employee’s role is to eﬀectively assis t the customer in the making choices, and fast service. The installer ’s role is to p erfectly mea- sure the dimensions and install the pr o duct with minim um amount erro r. The vendor’s role is to provide the requested pro duct deliv e red on an agreed date. The diﬀerent organiza tional units inv olved in this pro ces s a re the des ign center, the vendor with whom the order has been placed, the customer and the s ub- contractor who ins talls the pro ducts. The design center ﬁts right in the middle of this hierarch y , and needs to or ganize so that the customer’s needs ar e b eing met. The employees need to co ordina te with v endor to place the order and see that it is delivered. They also nee d to co or dinate with the installer so that the measurements and installations are done accor ding the customer’s conv enience and satis fa ction. 1.2.1 Organizational Chart The ab ov e ﬁgure shows the org anizational chart o f the all the p erso ns inv olved at the desig n center. Each arrow shows a relation of “ approaches to” fro m one per son to the other. An “ approach to” relation would mea n either information exchange or a need to consult. A unidirectional ar row mea ns one way a ccess and a bi-directional arrow implies b oth wa y approa ches. F or e.g., designer approaches the v endo r fo r placing an or de r in case of an ex- sto ck custom o rdered pro duct, with the information of type and quantit y . On the o ther hand vendor gets back to the designer with the informatio n on av ailability (the pr o duct might hav e b een discontin ued) and the delivery date of the pro duct. 7 Figure 1 .2: Organization Cha rt The diagra m do es not illustrate the “kind o f ” data or re la tionship b etw een the per sons. It merely depicts the need for exchange of information b etw een diﬀerent p ers ons inv olved at the design center. The org anization chart diﬀer e n- tiates b etw een p eo ple who are in outside the domain by representing them with a dotted line. 1.2.2 P erformer Roles 1. Designer : The designer is the pr ima ry user o f the automation system at the design center. Designer is resp onsible for helping c ustomer in pro duct selection, con tacting mea surer for taking the measurements, contacting vendor in ca se of custo m or der, main tenance of instock pr o ducts at the design a nd co n ta cting system analyst. 2. Cus tomer: Customer is the person who buys o r ha s in tent to buy a pro duct from the design cen ter. 3. Measurer : Measur er is the p erso n who takes measur emen ts o f the s ite and rep orts it back to the designer . He schedules his v is it to the site with customer upon req uest from the desig ner. 4. Installer : Insta ller is the per son who installs the pro duct at the site. He schedules his visit to the site with customer and is responsible for the installation of the pro duct. 8 5. System Analyst: Sys tem Analyst is resp onsible for maintaining the auto- mated system at Ho me Dep ot and ha ndling all technical pro blems within the s to re for the system. 6. Inv entory Mana ger: Inv entory manager deals with the in ven tor y related issues like or dering the pro ducts when they reach the critical level, main- taining corr ect level of sto ck for insto ck pro ducts base d on pr evious year’s sales and receiv ing the go o ds from the vendor for insto ck pro duct. 3 7. V endor: V endor is a contact in a co mpany that supplies pro duct for the design center. The desig ner cont a cts vendor for a custom order ed pro duct. 8. Cashier: Cas hier is the p er s on who accepts paymen ts from the customer against the pr o duct or services o rdered by the customer. 9. Customer Service Representativ e: Customer serv ic e repre sentativ e (CSR) is r esp onsible for int e r facing with custo mer for return of pro ducts and any other sp ecial requests by the custo mer. After the product is sold CSR is most preferred contact for the c us tomer r e g arding their previous o rders. Current Level of Requirements Do cumentation Man y of the activities that o ccur at the Design Center are pro cesse s a re generic to any r etail stor e . The existing s y stems do hav e do c umentation, which ar e listed below: 1. The training guide whic h not only helps new users get acquain ted with system but gives a list of the functionality provided by the existing system. The user s do ha ve knowledge of the ma jor r equirements that are needed by the a pplication. 2. There is another w ay o f ga ining the re q uirements by r everse eng ineering the ex isting applica tio n. 3. The r e a re also a n umber of manufacturer’s catalog s for e x ample Levlor , Decorama, Luxaﬂex etc whic h give informa tion about their pro ducts re- garding price, t yp e, dimensions, qualit y etc. T hes e catalogs give some selection pr o cess r equirements for the a pplication. 4. There also exist for ms and templates that are currently being used at Home Depot, which also for m a part of our requirement do cumentation. One o f the for m or template helps the design center employees in co llecting all the details from the customer , whic h helps them in placing the order without any mistakes. There exists another form known as the mar ked down for m that is used fo r rec o rding the da ma ged go o ds . The requirements th us consist of s o me exis ting do cumentation, r everse en- gineering the curre nt application and pap er based forms and templates. 3 In ven tory m anagement and r elated issues are out of the scope of the domain. Ho wev er, the i n terface to i n v en tory is part of the automation system. 9 1.3 Motiv ations for formal Systems Engineer- ing approac h and arc h itecture dev elopmen t The motiv atio n b ehind req uir ement s engineering is to get a clearer view of systems behavior with environmen ta l changes. Knowledge engineering is needed to formally understand each task p erfor med a nd its ‘wh y ’ in the system. Each phase of the for mal system engineering process has its o wn motiv ation factor, which has b een pr esented, in the following sections. 1.3.1 Motiv ations for Kno wledge A c quisition: 1. A t the design center the pe o ple involv ed in executing the sy stem manually bes t understand detailed working of the design center. Sessions of knowl- edge a cquisition a r e necessar y to unders tand the details of pr e sent day working o f the domain. Suc h an interaction is needed to ha ve b etter un- derstanding of the domain a nd its int e raction b etw een diﬀerent subsystem such as order tracking and billing. F o r example, the diﬀer e n t pr o cesses follow ed in custom order a nd in-s to ck or der of a n item. 2. Knowledge Acquisition sessions are needed to under stand the present le vel of automatio n a t Design Center. Parts of the domain are alre ady auto- mated like the inv entory manag e men t and c us tomer information tra cking system. There is als o a need to understand what kind of improv ements and changes can b e made for a b etter-integrated design cent e r automatio n system. 3. Knowledge exp erts of this doma in shall b e o ne of the end us ers of the sys- tem. Ther efore, the requir emen ts of the system will b e b etter understo o d in Knowledge Acquisition ses sions. This will help in setting b oundaries around the problem and the s olution spa ce and g ather the end user view- po in t o f the system. 4. T o under stand the interface and in terr elation betw een each subsystem (in ventory manage men t, custo mer service) such as order ing the items from vendor and mar kdowns of the items lost, stolen or da maged. 1.3.2 Motiv ation for Domain Mo deling: 1. It assesses crucial timing and resour ce requirements for subsystems such as delivery . The timing requirements being impo rtant fo r stor e ’s proﬁtability and customer s a tisfaction. 2. T o formally mo del the relationship b etw een the subsystems and entities such a s customer, vendor, in ventory , measurements call and delivery of go o ds for this domain. 10 1.3.3 Motiv ation for Reference Arc hit ec t ure: 1. Refer ence Arc hitecture serves as a blueprint to the developers of the automation system of design center. It deﬁnes the domain space a nd helps in der iv ing the implementation speciﬁc solution for Des ig n Center at Home Dep ot. 2. The reference Architecture is crucia l for system maintenance b ecaus e it serves as a map to the implementation and is easier to refer for a s y stem developer who is joining new or mo difying the system after it ha s b een implemen ted. It can b e referred to, for new pro duct additions at the design center of Ho me Dep ot. Suc h a dditio n could be the part of maintenance and/or future system upgr a de. 3. Reference Architecture for design center at Home Dep ot can b e used for other do main implemen tations, say design cent ers at JCPenn y , Sea rs and Low e’s. 11 Chapter 2 Kno wledge Acqui s ition 2.1 Ov erview Our team uses SEP A to disting uis h the functiona l requir ement s and implemen- tation sp eciﬁc requirements in or der to develop the application for the designer at Home Dep ot with co n tinuous ev aluatio n and tra ctability . This is obtained by: 1. Identifying the ma jor activities within the Design cen ter throug h in ter- active o bserv ation. 2. Un derstanding the cur rent level of automation and T echnology used at the Desig n ce nter 3. Creating an obser v a tion table for tra ctability . 4. F iltering the colle c ted data to generate sp eciﬁc scenar ios to develop a data dr iven sys tem 5. Planning KA sessions on the g enerated scenario s using analys is tec hniq ues such as 6. W ork pro cess Analy sis - T o conduct o bserv ation sessio n to identify the high le vel tasks accomplished b y the p erfor mers in the scop e of a given scenario. 7. T ask Analysis - T o inv estigate and classify tasks inv olved in carr ying out a par ticula r job function iden tiﬁed in a spec iﬁc scena r io and ev aluate it against the mental mo del 8. T emp ora l Analysis - T o identif y the duration, frequency and critical fac- tors to p erform a particular tas k . 12 9. Conceptual Analysis - T o understa nd the deﬁnition and relationship of one o b ject to another. It is also used to characterize attributes of reso urces and relationships b etw een data elements. 10. Scheduling KA session ba s ed on domain exp ert av a ilability , adequate time for the sessio n and the type of KA technique us e d. 11. Co nﬁrming session (date, time a nd place) with the domain exp e rt(Designer) to conduct formal Interview. 12. Mo tiv ate the domain exp ert to pr ovide input for the KA session, in order to design a system tha t helps the exp ert to do job b etter. 13. Providing a sample memo to the domain exp ert befor e the KA session, so that exp ert has thought a b out the scenario prio r to the meeting . 14. Analyzing and do cument ing the pr o cess using ﬂows a nd templates. 15. Reviewing the domain men ta l mo del and the r equirements outlined. 16. Constant ev alua tio n in each pha se for completeness and accur acy . 2.1.1 Observ ation w orksheet to do cumen t act ivit y Performer na me: Julie, Senior Designer Activit y Lo cation Information Resources Key P roblems Customer lik es to buy a prod- uct from de- sign center Design center at Home De- po t, Round Ro ck Customer can access the in- formation fr om web or from the catalogs W eb and catalogs Not enough informa- tion provided Customer se- lects wallpap er in sto ck or custom or ders Design center at Home De- po t, Round Ro ck Information to the customer is provided by the catalogs and the designer Catalogs and De- signer The pr o duct selected is out of sto ck or the design center do e s not deal with the vendor anymore Designer needs to co- ordinate with the vendor to get the de- tails(features) on the selected item Design center at Home De- po t, Round Ro ck Information to des ig ner is provided by the vendor vendor Designer may not b e able to reach the ven- dor to provide the de- tails on the pro duct useful to the customer 13 Customer or - ders 5 rolls of wallpaper Design center at Home De- po t, Round Ro ck Customer provides in- formation to designer 5 rolls of wallpaper 5 rolls of wallpape r do not exist. System is not upda ted to display the actual stock or the pro duct was no t or- dered or de lay in prod- uct delivery or unex- pec ted need for the pro duct Designer orders the pro duct us ing the mobile cart Design center at Home De- po t, Round Ro ck Designer ent e r s the infor mation in the mobile cart Mobile cart The system displays 5 r olls of wallpaper but only 4 exists, the 5th roll is torn Design center at Home De- po t, Round Ro ck Designer marks down the pro d- uct to 0 and en- ters the informa- tion o n a mark up and marked down shee t The mark up/mark down sheet Can lose the sheet and hence the infor ma tion Customer walks in and decides to buy the insto ck blinds at the design center Design center at Home De- po t, Round Ro ck Designer gives an option of diﬀerent t yp es of blinds and the pr ice ra ng e Catalog and t he Designer Customer selects the Insto ck blind and decides to buy it Design center at Home De- po t, Round Ro ck Designer enter the customer information and generates a n inv oice with the selected pro d- ucts a nd rela ted SKU# Dos pro- gram to ent er cus- tomer informa- tion and generate an inv oice Customer pays the bill and is satisﬁed with the pr o duct Design center at Home De- po t, Round Ro ck Customer pays the bill Customer , des igner and t he front d esk employ ee who is in charge of the bill 14 The mobile cart prog ram upda tes the ent ry for the pro duct sold in 24ho urs Design center at Home De- po t, Round Ro ck How Infor- mation g ets upda ted in the system and mobile car t Mobile cart Customer decides on custom – order blinds Design center at Home De- po t, Round Ro ck Designer pro- vides informa- tion o n diﬀerent t yp e s of blinds av ailable and price range on the v ario us kinds of blinds Catalog, customer and de- signer Designer takes the informa- tion r egarding the measure - men ts o f the windows to generate an exact price range Design center at Home De- po t, Round Ro ck Price range v a r ies on diﬀer- ent type o f blind and o n diﬀerent vendors a nd every type of blind has sepa- rate feature for which pricing is diﬀerent V endo r s catalog and Designer and fro n t end DOS application The vendors informa - tion is not av ailable on the system, the designer has to go through all the cat- alogs a nd note down the diﬀerent features manually , the desig ner may forg et to add a option or may enter certain data incor rectly The designer generates a inv oice for customer t o pay the bill Design center at Home De- po t, Round Ro ck Customer pays the bill Customer, designer and t he front d esk employ ee Customer requests for measurement and installa- tion Design center at Home De- po t, Round Ro ck The designer co-ordina tes with the p er- son incharge of measure- men t/installation and schedules an appoint- men t with the customer Customer, designer, Person incharge of mea - surement installation No prop er co- ordination 15 The designer sends the custom order requirements to the v endor and notes the delivery date Design center at Home De- po t, Round Ro ck Customer re- quirements send to the vendor Designer,vendor T a ble 2.1: Observ atio n W orksheet The ab ov e table was used to do c ument the activities obse r ved at the design center. This was obtained b y our frequent v is its to the design center and also observing the desig ner pe rforming these activities . The table also contains the resource s used and the problems encountered at the center. Ba sed on these activities we gener ated the scenario s for this domain. 2.2 KA T ea m Organization The KA team co mprises of : 1. Ansh uma n Sinha 2. Haritha Nandela 3. Vija ya B alakrishna Each member of the team has equal res po nsibilit y and has s imilar roles as the other team member. The r esp onsibilities for the Knowledge Acquisitio n were shar e d a mong the team member on the ba sis of scenar ios. Each team mem be r p erformed the knowledge a cquisition and the do cumented the rep ort for the sp eciﬁc scenario . 2.3 Scenarios driving KA eﬀo rt The sce na rios for the doma in ar e genera ted based o n the Interactive Obser v atio n conducted at the design center. Refer to ta ble-1 for Observ atio n w orksheet. There ar e many a ctivities that tak es place in the des ign center, we a re mainly fo cussed on the c us tomer interaction with the design c e nter. They ar e 1. Pro duct s e lection by the customer that is insto ck or custom-o rder 2. Customer r equesting q uote to make a decision on the pr o duct 16 3. Customer buy ing the pro duct 4. When custo mer retur ns the pro ducts 5. The proces s used by the design center if the customer requests for mea - surement a nd insta llation Based on the a bove topics our scena rios are genera ted a nd they tra verse through the doma in mental mo de l. These scena rios a re also ge ne r ated based on the co mplexity of each ta sk. The s cenarios dr iving KA eﬀort a re 1. Customer buys In sto ck pr o duct (Blind or wallpaper) 2. Customer s elects pro duct from catalog a nd r e quests for a q uote 3. Customer buys a custom o rder pro duct 4. Customer r equests for measur e men t, installation and delivery . 5. Return o f pr o ducts by customer 2.3.1 Scenario Descriptions Scenarios describ ed b elow represents the activities the p erfor mer s at Home De- po t. E ach scenario fo cuses on certa in asp ects of the domain thus generating requirements fo r the pro ject. Scenarios describ ed below represents the activi- ties of the p erformers at Home Dep ot. Each scenario fo cuses o n certa in asp ects of the domain thus g enerating req uirements for the pro ject. 2.3.2 Scenario #1 Customer buys In sto ck pro duct Scenario Description: J ane has just b o ught a new house r ecently and she would like to have window trea tments. Jane new home needs blinds for 18 windows. She needs to ﬁx the windows a s so o n as p ossible b ecause she is having a tenant come in next week. Jane feels that white color aluminum blinds would suit her home but she would like to conﬁrm her idea. She tak es the measurements for the all the windows and walks into Home Depot design cen ter and approaches the desig ne r . Jane gives details regarding her home and her idea s to Julie the designer. The desig ners tak es Jane to the display where all of blinds that are in sto c k are on display a nd asks her lo o k at them and decide on the type she likes. Jane ponder s o ver the details and dec ides that ma y be a luminum blinds would be right for her needs a nd that cream co lo r would match the wallpap er and would lo ok b etter. They lo ok at a couple of choices that a re in sto ck taking int o a ccount the price and other fea tur es like colo r , style etc. Jane ag rees tha t 17 white color 1-inc h blinds w ould suit her purpose b etter. Julie lo oks up in the inv entory and sees that they have Bali blinds in standa rd s ize in sto ck. The blinds are not av ailable for some window size s and Julie o ﬀers to cut the blinds for Ja ne. After Julie cuts the blinds, Ja ne lo ads up her cart with all the blinds walks to the chec k out desk and the c le rk scans in all the items. Jane pays for the items b ought and lo a ds the items int o her vehicle and takes them home. Scenario Ob jective: T o understand the pro ces s in volved in customer se le c ting an in sto ck pro duct and also buying the pro duct. Justiﬁcation: In this particular scena rio we are b e consider ing the customer buying a pro duct be it either wallpap er o r Blinds that is currently in the ware- house. The process of dealing with customers who buy in sto ck pro duct is diﬀerent from a custom order pro duct. Therefore there is a possibility o f iden- tifying additiona l tasks no t identiﬁed in other scenar ios. In this scena rio we will cover the following work pro cesses o f the mental mo del namely the Select pro duct, Pr o cess Order , Buy Pro duct. Refer to ﬁgure 2.1 Figure-2.1 -Depicts the mental mo del for s cenario#1 2.3.3 Scenario #2 Customer requests quotes on the pro d- uct selected from catalog Scenario Description: Cynthia walks in to design ce n ter , Home Depo t with measurements for all her windows in her house to buy blinds. She has no prior selection o n the type of pro duct. The designer presents her the v ar ious choices av ailable a t the design center, like the Insto ck pr o ducts o r the catalog pro ducts. She explains to the customer the adv a nt ages of each type of pr o duct, the v arious options the vendor can provide, the diﬀeren t t yp e of v endor s av ailable. Bas ed on this infor ma tion she sele cts the pr o duct from cata log and req uests a quote from the designer. Ba s ed o n the quote obtained she makes her decision whether to buy or not to buy the pro duct. Scenario O b jectives: T o understand the pro cess inv olved in customer ap- proaching the desig ner after selecting the pro duct from c a talog, desig ner gener - ating quo te based on customer measur e ments for the type of pro duct selected. Justiﬁcation: This scena rio traverse throug h one of the main tasks “Gener- ate Quote” of the do main men ta l mo del where the initiating even t is customer selecting a pr o duct fro m catalog and approaching the des ig ner for quo tes. The designer understands the customer requirements and g enerates quotes as r e- quested. Ba sed o n these quo tes the customer ma kes a decision to buy the pro d- uct. The text repr esented in bo ld in the domain mental mo del b elow, ﬁgure-2.2 represents this par t of the scenar io. Figure-2.2 -Depicts the do main mental mo del for scenario#2 18 2.3.4 Scenario #3 Customer buying the custom order pro duct Scenario D e s cription: Cynthia g ets a quote from the designer ba sed o n the pro ducts she selected from the ca ta log. She can now make a decision based o n this quote. She takes ab out tw o days, a nd ﬁnally ma kes her decision to buy the pro duct and returns to the desig n center at Home Dep ot to buy it. The designer conﬁr ms with her tha t she has the right measuremen ts and asks her where she wan ts the pro duct to b e deliv er ed. If Cynthia has not taken cor rect measurements, the designer asks her to re-measur e and gives enough information how to do it. Cynth ia w a nts it to b e delivered at Home depot and she would like to c ollect it fr o m home depot. Designer generates an invoice based on her selection and tells her a bo ut the delivery date, a nd Cyn thia pays for the pr o duct at the cash r egistry . Once the pro duct ar rives at the home dep ot, the designer calls Cy n thia and a sks her to collect he r pro duct. Scenario O b jectives: T o understand the pro cess inv olved in customer ap- proaching the designer to buy the pro duct and the designer genera ting inv oic e for the customer and the customer paying the bill. Justiﬁcation: This scenario trav erse through o ne o f the main tasks “Buy pro duct” of the domain mental mo del where the initiating even t is customer receiving a quo te from the designer a nd making a decision to buy the pro duct. The designer ge ne r ates the inv oice f or the cus tomer and the customer pays for the bill. The text represe n ted in bo ld in the domain men tal mo del below, ﬁgure-2.3 b elow re pr esents this part o f the scenario. Figure-2.3 -Depicts the Doma in mental mo del for s cenario#3 2.3.5 Scenario #4 Customer selects an insto ck product and requests for measuremen ts, installation and de- liv ery . Scenario Descr iption: Itsy comes to Home Dep ot Design Ce n ter to buy regula r blinds for her new home in T own Lake. With the help of the desig ner she chooses one of the blinds that is in sto ck fo r living ro om of the hous e . She is not sure of the measure ments of windows in living ro om and w a nt s the mea surements to b e taken b y the meas urer of the Design Cen ter . She chooses a no ther t yp e of blinds fr om the catalog, fo r a ll other ro oms on gro und ﬂo or of her duplex house. The measurements of a ll the windo ws in those ro o ms are also to b e taken b y the mea surer. Itsy will b e out of town fo r the next couple o f days a nd wan ts the measurer to come only after she is back in town. She would prefer the following W e dnes day after 6:0 0 p.m. after she gets back fro m work. W hen the measur ement s are taken she will come ba ck to Home Dep ot and pay for the blinds. She also wan ts the blinds to be installed by the installer. The installer should contact her to make a n app ointmen t and agree on suitable time and 19 date. She would like the installa tion to b e done in the week end. The blinds are to o big to ﬁt in the car a nd there fore she needs it to b e delivered by Home Depo t. She is ready to pay extra for the delivery and installation. Scenario Ob jective: T o understand the pro cess inv o lved in ta king meas ur e- men ts, installa tio n and delivery o f items b ought by the customer. Justiﬁcation: The de s ign center oﬀers the s e r vice of measuring the site b y a suitable p erson for its customers. T o achiev e c ustomer satisfa ction and to deliver the righ t quantit y of pro duct in correct size, taking measurements b ecomes an imp or tant activity for the design center and needs to b e detailed. This scenario fo cuses on this a ctivity to ga ther all requirements for this pr o cess. The scenario also covers the installation a nd delivery o f an in sto ck pro duct. T aking measurements, delivery and Ins tallations are not dep endent on the v a riables like the make o f the pro duct and its s tock av aila bilit y . This scena rio pictures a po sition that happ ens quite regularly at the design cent er and the automation system will need to s e r vice to such r equests. Refer to ﬁgure 2.4 b elow. Figure-2.4 -Depicts the Doma in mental mo del for s cenario#4 2.3.6 Scenario #5 Customer returns a pro duct. Scenario Description : T eene comes to Home Depot in Round Ro ck customer service desk with a set of wallpaper s that was purchased by her husband for their new house in Cany on Oa ks. W ean, her husband, b ought the wallpaper a t the Home Dep ot on Resear ch Blvd last mont h. She wan ts to return a part of the purchase made by W ea n and buy another s et of wallpapers fo r their living ro o m. If she doesn’t ﬁnd something that she likes at Home Depot design center, she will try o ther plac e s in town. The customer ser vice desk accepts the returned items a nd forwards it to the designer at the design center. Scenario O b jective: T o understand the pro cess inv olved in customer return- ing an insto ck pro duct. Justiﬁcation: The most imp ortant goal at Design Cen ter is to sell the pro duct that is suita ble to customer’s needs and the p er fo rmers constantly try to achiev e this b y impr oving the system. Home Dep ot not only tries to achiev e this but also ha s the ﬂexibility of accepting the return of items when the pro duct do es not meet c us tomer’s sp eciﬁcatio n and s atisfaction. This scenario tries to trace how the pro duct return is handled with resp ect to des ign cen ter. Spea king in terms of the eﬀect of return o f go o ds on billing and inv ent ory , the sub ject is out of the scop e of doma in. How e ver, the desig ner at the design center is resp onsible for accounting the returned g o o ds and making a decision on returned pro duct in case o f in sto ck pro duct. This scenario tries to tr a ce the pro cess, which happ ens at the design center a fter the item has b een handed ov er to the designer by the front desk . Refer to ﬁg ure 2.5 b elow Figure-2.5 -Depicts the Doma in mental mo del for s cenario#5 20 2.4 KA Plan S. No. KA Session Title Doma in Exp ert Date/ Duration KA T echnique Scena rio 1. Customer buys In- sto ck pr o duct Jimmy 2/ 2 8/00, 5.00p.m 1.5 hr s. W o r k pro ce s s Analysis, T ask Analysis #1 2. Customer buys In sto ck pr o duct Jimmy 3/ 0 4/00, 11.00a .m 1.5 hr s. T e mp or al Anal- ysis, Conceptual Analysis #1 3. Customer selects a custom-order pro d- uct Julie 3/ 06/00 6.00p.m 1.5 hr s. W o r k pro ce s s Analysis, T ask analysis, T em- po ral Analysis, Conceptual Analysis #2 4. Customer requests quote on the pro d- uct sele c ted from catalog Julie 2/ 29/00 5.00p.m 1.0 hr s. W o r k pro ce s s Analysis #2 5. Customer requests quote on the pro d- uct sele c ted from catalog Julie 3/ 2/00 10.00a .m 1.0 hr s. T a sk Analys is #2 6. Customer requests quote on the pro d- uct sele c ted from catalog Julie 3/ 6/00 5.00p.m 1.0 hr s. T e mp or al Anal- ysis, Conceptual Analysis #2 7. Customer buys the Custom order pro d- uct Julie 3/ 14/00 10.00a .m 1.15hrs W o r k pro ce s s analysis, T as k Analysis #3 8. Customer buys the Custom order pro d- uct Julie 3/ 20/20 00 6.00p.m 1.0 hr . T e mp or al Anal- ysis, Conceptual Analysis #3 9 Customer requests for mea s urements. Lisa 3/04/ 00 7.00 p.m 1.0 hr W o r k pro ce s s analysis, T as k Analysis, T em- po ral Ana ly sis #4 21 10 Customer requests for deliv er y , instal- lation and return of pro ducts. Lisa 3/18/ 00 5.00 p.m 1.5 hr s. W o r k pro ce s s analysis, T as k Analysis, Con- ceptual Analysis #4, #5 T a ble 2 .2: Knowledge Acquisition Plan 22 Chapter 3 KA Sess ion Rep orts 3.1 KA Session # 1 3.1.1 Session I nformation Scenario E mploy ed: Scenario #1 : Customer buys an in s to ck pro duct. Doma in Exp ert: Jimmy Rice Title: Senior Designer E mploy er: Home Dep ot Ba ckground Information: Jimmy has be e n working at HD for the las t three years and he is very exp e rienced in the design center business . He helps the custo mer select pro duct, pro cesses the order for them and makes their visit to HD a pleasur able one. He is also resp onsible for all the num b er crunching that takes place a t the design c e n ter . He has a couple of a sso ciates who a ssist him and who lo ok up to him for guidance. Knowledge Engine e r: Haritha K Na ndela Session Time and Date: 5.0 0 p.m. 02 /28/0 0 Session Lo ca tion: T raining r o om a t Home Dep ot, South Austin Session Purp ose /Goal: T o a scertain enough information for the work pr o cess Select Pro duct, P ro cess Order and Buy Pr o duct for an in sto ck pro duct so that the diﬀerent tasks and sub tasks in the pro ces s ca n be iden tiﬁed. 3.1.2 Session Input 1. The KE needs to r ead the textbo ok titled Use r-Centered requirements and class no tes so as to pr e pare herself for the diﬀere n t type of questions to be asked. 2. Before the sessio n the KE needs to determine the techniques she is going to be using for the coming session. 3. The KE a lso ha s to prepare a ques tio nnaire that would help her in a na- lyzing the work pro ce s s and task s. 23 4. The KE needs to fax a memo to the DE. Pleas e refer Fig 2.7 on the next page. 5. The KE also has to a rrange for spe ciﬁc time and pla c e to meet the Doma in Exp ert. 3.1.3 Session Output 1. A n o verall understanding of the work pro ce ss Select Pr o duct, Pr o cess Order and Buy Pr o duct and g ain information regar ding the reso urces a nd that da ta requir ed for those pr o cesses. 2. Ident ify diﬀere n t sub tasks a nd tas ks for the work pro c esses. KA T echnique(s): W or k Pro ce s s Analys is , T as k Analysis. Type of Knowledge: P r o cedural knowledge 3.1.4 Planned KA Questions 1. Is scenario #1 a v alid scena rio? 2. Who a re p eople inv olved in this pro cess? 3. W ould you like to change anything in pro cess cur rently av a ilable? 4. F o r the given scenario w ha t do you after cus to mer walks in? 5. How do you extract the information fro m the customer to help him in choosing the pro duct? 6. How do you appro ach a p erson trying to buy wallpap er? 7. Are there diﬀerent gr ades of wallpap er? What ar e they? 8. If the customer k nows the co lor of wallpap er that he likes how you dir e ct them a r e the wallpaper b o ok s sor ted by color ? 9. What do you base the selection of wallpap er for a customer who ha s come to buy in sto ck? 10. Ho w do es the c ustomer decide how m uch wallpap er he needs? 11. F or an in sto ck pro duct after the selectio n of the pro duct what’s next? 12. If the cus to mer has windows requirements that you do not have in stock for example if he needs 15 a nd you have o nly ten of them wha t do you do? 13. Do you gener ate any quotes for in sto ck blinds? 24 14. Do you gener ate any in voices for in sto ck blinds? 15. F or in sto ck pro duct do you tell the customer how to install the blinds and do they need accesso r ies? 16. Is there a diﬀerence in an individua l homebuyer and a co mmercial buyer? Session Rep or t Date: 03 /11/0 0 T o: Jimmy R ic e Senior Designer Home Dep ot F ro m: Hari t ha K Nandela Hel lo Jimmy, I was wondering if it would b e okay with you to me et on 02/28 /00 at 5.00 p.m. for m ay b e an hour or so. I would like to discuss with you ab out the fol lowing pr o c ess at home dep ot. I wil l b e c overing the or dering of a pr o duct that alr e ady available in sto ck. I wil l b e doing it for blinds as wel l as wal lp ap er. If you have any qu estions ple ase c ontact me at this n umb er 512-723 -8445. Thanks Haritha 1. Is scena rio #1 a v alid scenario ? A: The pro cess is correct but for scenario #1, you might nee d to change a c o uple things like the customer would buy aluminum blinds instead of w o o den. The reason why the customer buys the in sto ck aluminum blind is b ecause they want something for short term and something that easily r eplaceable. They might b e landowner as oppo sed to homeowner. As a general r ule of thum b homeowners who int e nd to stay for a long time would invest more time and money in buying blinds but s omeone loo king for co mmercial rea sons would go for something cheap a nd fas t. 2. Who a re pe ople inv olved in this pro cess? A: Deﬁnitely the ho meowner, designer, vendor, installer and the cashier. 3. W ould you like to c hang e a nything in pro cess currently av aila ble? A: Something that I would like to change is that if it were p o ssible to link the diﬀerent Home Depo ts so that I could acces s the information regarding inv entory at diﬀerent lo catio ns. Also if I could ac c ess the data o f diﬀerent vendors so that I can v er ify that they will b e able deliv er the requests I make b efor e I give the c ustomer a n y kind o f estimates etc. 4. F o r the given scena rio what do you after the cus to mer w a lks in? A: When customer walks in I try and under s tand fro m the informa tion wha t her goal of c oming to home dep ot is. After s he tells me that she is planning on installing blinds for her home which she is going to rent out to tena n ts, I realize that she is lo oking for so mething cheap and fa s t. I direct her to the display where we have diﬀerent kinds of blinds on display . Since she would like something she c o uld pick up and walk out, I’d suggest her to lo ok at aluminum blinds because w e hav e a lo t of sizes and in that t yp e a nd she 25 might ﬁnd all windows that she needs. T he n she’d s elect the color a nd then dep ending on the window s izes I’d cut them the blinds if r equired. 5. How do you extract the information fro m the customer to help him in choosing the pro duct? A: W e call that pro cess qua lifying the customer. What it means is we need to know who, what, w he n, where and why . Who is the decision-maker, it’s not go o d if your friend comes and tells that she wan t blinds for y our ho use I’d prefer if it w ere you who comes in. When is the when are you go ing to buy the pro duct is it going to be one year from now or is go ing to be next week. Where is where ar e you and where you would like to insta ll the pro duct. Do you wan t it in Montana s o do I need to refer you to another Home Dep ot or can I help you. Wha t is wha t are you trying to do are you? Are you r emo deling the house ? or you wan t new windows for the whole ho use etc. Why is basica lly if know what your g o als are for your home pro ject. So this is the kind of information I tr y to gather from the customer and if can get the answers for all these questions then I’m in very go o d p osition to help the customer. 6. So what is the approach when so meone come in to buy in stock blinds? A: W ell, ﬁr st I’d like to know their pur po se a nd then I’d direct them to the display from where they can decide what kind o f blinds they wan t. F o r exa mple horizo n ta l, vertical etc. 7. How do you a ppr oach a p erson trying to buy wallpap er? A: Some of the customers do hav e an idea what kind of wallpaper they want if they do not I ask them to w alk ar ound and s ee the diﬀerent types in sto ck and see if they like a n y of them. If they do not ﬁnd anything int eresting then I a sk them to lo ok in to the wallpap er b o oks. Where there are diﬀerent sections for contemporary , kitchen, bathr o om etc., which might aid them in cho osing the pro duct. 8. Is everything that is in s to c k on the shelf or is it somewhere in the w are - house? A: Everything is normally on the shelf other wise it might b e in the overhead but everything is in the stor e. 9. If the customer k nows the co lor of wallpap er that he likes how you dir e ct them are the wallpaper b o oks sorted by color ? A: No the wallpaper b o oks are not sorted by co lor but you ﬁnd the co lor you wan t in every b o ok. The ﬁrst thing the customer needs to decide on is the style of wallpap er that are contemporar y , traditional, kitc hen and bath. Then the b o oks are divided based on the section like for ex a mple if the customer wan ts cowboy wallpap er. I’d direct them to the country sectio n a nd there they would ﬁnd the cowboy wallpaper in b eige color. 10. Are there diﬀerent gra des of wallpap er? What ar e they? A: Y es, there are diﬀerent gr ades o f wallpaper. They are Solid vinyl, Vin y l coated and Embossed. The solid vinyl is very dura ble a nd can b e used in k itc hen. The vin yl coated is has a thin la yer o f vin yl ov er the pap er and the la st 26 one is the embossed wallpap er that is w allpa p er with puﬀy impre s sions. The embossed in very exp ensive is not very durable. W e sugg est the so lid vinyl b ecause they ar e more durable. 11. F or an in sto ck pro duct after the selection of the pro duct what’s next? A: After the customer selects the pr o duct then I a sk them if they have made a color selection. After they have decided I will go see if the list of window sizes they need are in sto ck. If the sizes do not match we can cut the blinds bas e d on the mea surements given by the customer. When do ne I lo ad them on to the cart and send to them to the cash register. 12. If the cus to mer has windows requirements that you do not have in stock for exa mple if he needs 1 5 and you hav e only ten of them what do you do? A: F or in sto ck pro duct we ca n cut the blinds in sto ck give them to the customer for the size he needs. W e ﬁr st can lo ok at the ov erhea d if its av ailable o r check if its av ailable at o ther stores or we can tell him when they a re g oing to in sto ck and as k him to come back a gain. They ca nnot pay a head and come back and collect them 13. Can the as so ciates help the customer with the in stock pro duct? A: Ab- solutely , the ﬁrst thing the asso ciates a re taught is to cut the mini blinds and know the diﬀerences b etw een the t y pes . The in sto ck is what we sell most but it’s not the place we make mor e money . 14. Ho w do you cut the blinds? Do you hav e any equipment? A: Y es we hav e a mini blind cutter that is given to us by B ali, we use it to cut the blinds. W e cut I-inch aluminum and vinyl and we cut only B ali brand and no other. 15. Do you generate any quo tes for in sto ck blinds? A: Normally I do not. The customer himself or herself walk ar ound lo ok up the prices and the sizes they need a nd make the quote. F o r example for Bali blinds the there is list on the shelf that says 30x6 0 is 1 7 dollars etc. Sometimes if the customer needs some help we help them gener ate a quo te o r do it fo r them. 16. Is there a n a dditio na l tas k if the pro duct is wallpape r ? A: There is a n additional task for wallpaper b ecause the note on the wallpap er says that just dip it in the water and stick it. W e recommend that they use the glue .W e hav e a whole w all lined with diﬀeren t glues. W e tell them ho w prime the wall, use the glue and hang the wallpape r . So I spend some time ex pla ining to the customer the who le pro cess. 17. Ho w do es the customer decide how muc h wallpaper he needs? A: W ell the easiest wa y to decide is calculate the square footag e for the w alls in the ro om and deduct the do ors and windows which would give an approximate square fo otage . Each roll has 5 6 square feet s o from that I calculate a nd tell them how many rolls they need to buy . 27 18. Do you genera te any inv o ices for in sto ck blinds? A: No we do not do any inv oice generation for in sto ck pr o ducts 19. Can the cus tomer ta ke ha nd s amples for the in s to ck blinds? What is the pro cess for wallpaper? A: No we do not give sample bo o ks to customer to take home. But the wallpap er is on the stands and the customer can tea r oﬀ a piece of the wallpap er and ta ke it home to see it is a go o d selectio n. 20. What do you bas e the s election of wallpaper for a customer who has come to buy in sto ck? A: The w allpa per for kitchen sells the fastest. B ut I’d ask them questions lik e the c o lor of the cabinets , light o r dark wallpap er, wall s pa ce etc. Then I’d direct them to the r ight sectio n. 21. What are the types o f blinds that you hav e in sto ck? A: In ho r izontal blinds I have 1-inch vinyl and a lumin um and 2-inch vinyl and in vertical I have diﬀerent styles in plastic. 22. F or in sto ck pro duct do you tell the customer how to install the blinds and do they ne e d acces s ories? A: Y es, I give them quick run down how to do it. T he pro duct has instr uc tio ns on how to install it so that will help them. They do not need any a ccessor ie s to install the blind they might need a drill but that’s abo ut it. 23. Is there a diﬀerence in an individua l homebuyer and a co mmercial buyer? A: Not really . The o nly thing I ca n think of is that the commercial cus- tomer mig h t hav e more windows to cut and pull oﬀ the shelf which might take some time as opp os e d to a homeowner who migh t need o nly so me windows. 24. What are diﬀeren t wa ys we can pay at home dep ot? A: At home depot you ca n by cash, chec k or card. Any kind of credit card is ok ay . 25. Ho w w ould you schedule the asso ciates? A: What I do is ma ke sure that a fresh as so ciate is no t left all alone. And that there is some s enior p e rson with her a ll the time. 3.2 KA Session # 2 3.2.1 Session Information Scenario E mploy ed: Scenario #1 : Customer buys an in s to ck pro duct. Doma in Exp ert: Jimmy Rice Title: Senior Designer E mploy er: Home Dep ot Ba ckground Information: Jimmy has be e n working at HD for the las t three years and he is very exp e rienced in the design center business . He helps the custo mer select pro duct, pro cesses the order for them and makes their visit to HD a pleasur able one. He is also resp onsible for all the num b er crunching that takes place a t the 28 design center. He has a couple of asso ciates who assist him and who lo o k up to him for guida nce. Knowledge E ng ineer: Ha ritha K Nandela Sess io n Time and Da te: 11.00 am 03/ 04/00 Session Lo cation: T raining ro om at Home Depo t, South Austin 3.2.2 Session Purp ose T o ascer tain the temp or al and conceptual information for the work proces s Select Pro duct, Pr o c ess O rder, Buy Pro duct for an in sto ck pro duct as to derive time line diag r ams and c o ncept mo dels. 3.2.3 Session Input 1. The KE needs to r ead the textbo ok titled Use r-Centered requirements and class no tes so as to pr e pare herself for the diﬀere n t type of questions to be asked. 2. Before the sessio n the KE needs to determine the techniques she is going to be using for the coming session. 3. The KE a lso ha s to prepare a ques tio nnaire that would help her in a na- lyzing the work pro ce s s and task s. 4. The K E faxes a memo to the DE. Plea se refer to ﬁg 2.8 5. The KE also has to a rrange for spe ciﬁc time and pla c e to meet the Doma in Exp ert. 3.2.4 Session Output 1. An ov erall understanding timing issues and concept r elated to the diﬀerent work pro cesses Select Pro duct, Pro cess O rder and Buy Pro duct. 2. Ident ify for e a ch task in the work pro cesses the pre and p ost conditio ns , task fre q uency and task duration. KA T echnique(s): T emp ora l Analysis, Conceptual Analysis. Type of Knowledge: Semantic knowledge Session Rep or t Date: 03 /12/0 0 T o: Jimmy R ic e Senior Designer Home Dep ot F ro m: Hari t ha K Nandela Hi Jimmy, I was wonde ring if i t would b e o kay with you to me et o n 03/ 04/00 at 11.00 am for may b e an hour. As b efor e I wil l b e c overing t he or dering of a pr o duct that al r e ady avai lable in sto ck. But this time I wil l c onc entr ate on 29 the timing issues and c onc erns and also the any other terms t hat ar e use d in this pr o c ess. If you have any qu estions ple ase c ontact me t he numb er is 512- 723-844 5. Thanks Haritha 1. How muc h time you sp end in the selectio n pro cess co mpared to the whole pro cess of buying the pro duct? A: The time depends if the customer wan ts a pro duct that in sto ck or if he wants a custom order. F or in sto ck pro duct I’d just walk them to the displa y and show the pro duct a nd let them select the product so I ma y b e s pend not mor e tha n one minute o r so. If they wan t some help in co lor selec tio n I might take some more time but otherwise I will sp end a lot time. Normally a customer who wan ts a n in sto ck pr o duct is not re a lly in ter ested in sp ending a lot of time a t HD. Compared to buying a sp ecial orde r pro duct that you will be living in for the res t of your life to buying an in s to c k pro duct the time sp ent is very less. 2. What is frequency of customers coming to buy? A: W e hav e customers coming in to buy a ll the time the freq uenc y is more during the weekend, evening. I no rmally hav e time at the end of the day to spend with the customer. 3. Of the customers co ming how many really buy a pro duct? A: There are diﬀerent customers coming into the design c e n ter a bo ut 50% ab out o f them hav e an idea o f what they a nd just wan t you to conﬁrm their dec ision and they buy the pro duct and are o n their way a nother 25 % want to you tell them how to install the pro duct and the re s t do not know anything ab out the pro duct they are a bo ut buy and need to hand held all the w ay throug h. 4. If a c ustomer do es buy a pr o duct what is the mos t sold pr o duct? A: The in sto ck blinds ar e the most s o ld how ever it is not the most mo ney ma king pro duct the sp ecialty orde r are the most money ma k ing sectio ns . 5. Do you get any sort of reward for making s ales? A: Not really . But at the end of the year all the num be r a re drawn up for a ll the sales a nd the returned go o ds that kind o f gives an idea how muc h money I ma de for the company . And gives an ov era ll pic tur e for the p erfo r mance r eview. 6. Whom do you co nsult for questions r egarding the pro duct? A: Nor mally I would cons ult the vendor for ques tio n I do not hav e answers. 7. What a re diﬀerent levels of exp ertise at design c e n ter ? A: Y es, Curr ently we hav e new a s so ciates w ho a w eek old. So we hav e diﬀerent levels. The levels are b eginner, mo derate and exp ert. A beg inner with six months of exp erience ca n s tart to help the customer with a lot of eﬃciency 8. How long do es it take to cut the blinds ? A: It g enerally dep ends on the nu m be r o f blinds. Nor mally it takes ab out 4 min utes p er blinds and that’s what w e tell the cus to mer. 30 9. Can the as s o ciates do the cutting of blinds? A: Y es they can. Normally they fr eeze up if the customer asks sp ecial o rder q uestions. But they ca n do other tas ks like cutting the blinds. 10. What is fr equency of cutting the blinds when a customer is buying an in sto ck pro duct? A: F o r in sto ck blinds we hav e many sizes in sto ck for the po pular brands . So may b e half the time we might need to cut the blinds. So 50% of the time the customer can just pic k the pro duct and walk to the ca s hier’s desk. This is b ecause for an in sto ck pro duct y o u will not get a p erfect ﬁt as it is with a custom order blind. 11. Before you cut the blinds do you as k the customer to pay ahead of time? A: No. After w e cut the blinds they take the pro duct to the fro n t desk. Sometimes w e get phone calls fro m a custo mer who has already b een to the desig n center and they a s k for the blinds to b e cut and ne ver co me and pic k them up. 12. Ho w long do es it take for a custom to pick, customize and buy an in sto ck pro duct? A: It do es no t take more than tw ent y min utes for the customer to s elect, buy and take the pro duct and walk aw ay with the out. Since the time to cut each blind is four min utes it really dep ends on the num b er of blinds that need to b e cut. The range might b e 20 min utes to a maximum of t wo hours . 13. If someone wants to buy in sto ck blinds what is the most cr itical infor- mation tha t the perso n needs to hav e? A: The most p opular t yp e for in sto ck is the aluminum blinds so if they hav e the measur emen ts then w e can help them decide and buy . 14. What kind of mea surements do you need to c ut the blinds A: The insto ck blinds can never b e a s go o d ﬁt as the custom order ones and that is why they are so cheap. But we can make adjustments with the in sto ck blinds so w e may hav e a leewa y of an inc h. So y o ur measure ments need to be plus or minus inch. 15. Where do you get data or informa tion to contact the vendor? A: W e hav e a price b o ok that has the sku num b er a nd it a lso has the vendor co n tact information. 16. What ar e the diﬀer en t accesso ries you ca n buy for wallpap er? Do you give discounts? A: F or the wallpaper you can buy the wallpape r itself and the bo rder and the fabric to match. Y e s w e do g ive dis c ount s o n the pr o duct. 31 3.3 KA Session # 3 3.3.1 Session Information Scenario Employ e d: Scenario #2: Customer se le cts a custom order pro duct. Domain Exp ert: Julie Title: Senior Desig ner E mploy er: Home Dep ot Ba ck- ground Information: She is working for the Home Depot Design cen ter for the past 4 years. She helps the cus tomer of Ho me Dep ot to make the right pro d- uct selection a nd generates quotes for them to make a decision on the pro duct. Knowledge Eng ineer: Haritha K Nandela Ses sion Time and Date: 6 .00 p.m. 03/06 /00 Session Lo cation: Home Dep ot Design Center, Round Ro ck, Austin. 3.3.2 Session Purp ose T o ascertain eno ug h information for the work pro cess Select P ro duct for a cus- tom order pro duct so that the diﬀeren t ta sks and s ub tasks in the pr o cess can be identiﬁed. 3.3.3 Session Input 1. The KE needs to r ead the textbo ok titled Use r-Centered requirements and class no tes so as to pr e pare herself for the diﬀere n t type of questions to be asked. 2. Before the sessio n the KE needs to determine the techniques she is going to be using for the coming session. 3. The KE a lso ha s to prepare a ques tio nnaire that would help her in a na- lyzing the work pro ce s s and task s. 4. The K E faxes a memo to the DE. Plea se refer to ﬁg 2.9 5. The KE also has to a rrange for spe ciﬁc time and pla c e to meet the Doma in Exp ert. 3.3.4 Session Output 1. An ov er all under standing of the work pro cess Select Pro duct for a custom order and gain information r egarding the reso urces and that data required for thos e pr o c esses. 2. Ident ify diﬀere n t sub tasks a nd tas ks for the work pro c ess. KA T echnique(s): W or k Pro ce s s Analys is , T as k Analysis. Type of Knowledge: P r o cedural knowledge, Semantic knowledge 32 3.3.5 Planned KA Questions 1. Could you elab orate for me the diﬀeren t steps inv olved in selecting pr o duct like blinds? 2. If the customer walking in do es not know anything ab out the blind u show him the diﬀer en t types but if ha s a n idea ab out what he wan ts then how do approach? 3. So what ar e the diﬀerent tasks you do when a customer co mes in and says that I’d like blinds? 4. Whic h are cr itica l factors that are imp ortant so as to b e successful in the selection pr o cess? 5. What r eally helps the customer in deciding from the diﬀer e nt types? 6. Are blinds sp eciﬁc to a par ticular ro om? 7. How do you make s ure the customer choice is cor rect? 8. How long do es it take to select a s p ecia l order pro duct? 9. When qua lifying the customer what is the kind o f infor mation you are trying to g e t? 10. If the customer has a sp eciﬁc request w ha t do you do? 11. When I make a selection what do? Do you contact the vendor for av a il- ability? 12. If the vendor has discontin ued the pro duct? What is your next step? 13. What kind o f details do you give the vendor when you asking him for av ailability? 14. Ho w do you contact the vendor? If you a re not able to r each the vendor what do you do? 15. Does the vendor hav e a n y delivery date s p eciﬁed? 16. Do you sugg est any a ccessorie s when the customer selects a blind? Session Rep or t Date: 03 /12/0 0 T o: Julie Senior Designer Home Dep ot F ro m : Hari t ha K Nandela Hi J immy, I was wondering if it would b e okay with you to me et on 03/ /00 at 6.00 p.m. F or this session I wil l b e c overing the sele ction pr o c ess for buying a custom or der pr o duct. I wil l c onc entr ate on the r e quir ements for the pr o c ess and the diﬀer ent tasks involve d in the pr o c ess. If you ha ve any qu estions ple ase c ontact me the numb er is 51 2-723-8445. 33 Thanks Haritha 1. Could you elab orate for me the diﬀeren t steps inv olved in selecting pr o duct like blinds? A: The ﬁrst thing I do when the cus to mer walks in is take them to the display a nd show the diﬀerent types av a ilable like the ce llula r shades, 2- inch a nd verticals. So I get them to ﬁrst decide o n the type of pro duct and then show them the diﬀerent bra nds av ailable in that particular type. Normally I s how them three br ands b ecause the customer gets c o nfused if you show them more than three brands. Then I tell them the adv a ntages, disadv antages, price, w ar ranties a nd diﬀerences. Finally the decision is dr iven by the price and the color, whichev er brand o ﬀers a color the customer wan ts a t reasona ble price that’s the pro duct they pick. 2. If the customer walking in do es not know anything ab out the blind u show him the diﬀer en t types but if ha s a n idea ab out what he wan ts then how do approach? A: In the b eginning when deciding on the type of blinds price is the impor tant factor that helps the customer decide. I pick out a standard size and g ive the cost for that s ize for the diﬀerent t y pes , which helps him decide. 3. What is the appro a ch you take when the customer walks in design center? A: First th ing we do is try and identify when the customer wan ts the window treatmen ts or wallpaper so that we can assis t him b etter. Also understand their pro ject whether it’s a r emo deling pro ject or a new hous e that they trying to get window treatmen ts for . Identif y how serious the customer in buying the pro duct. If they are not sure I try and dir ect them to diﬀer e n t b o ok s from where they can get so me idea s. I s ug gest them some ideas, give them the o ptions they hav e av ailable. F or example I had a customer yesterda y who star ted building I realized that they would not need the window treatmen t un til next year so I just show ed them t y p es and asked them to observe window tr e atment s whenever they go anywhere. 4. So what ar e the diﬀerent tasks you do when a customer co mes in and says that I’d like blinds? A: I’d ﬁrst take them and show them diﬀerent types and giv e them the hand samples. So the ﬁrst factor they need to decide is the t yp e because I do not wan t to spend tw o hours and then ﬁnd out that then customer would like another type. So qualifying the customer is the most imp ortant thing. Be c ause I’d lik e the customer to decide what they want or tak e time to decide what they wan t. So a fter the cus to mer decides the t yp e and then I’d take them to the sample b o oks show them the sample b o oks in diﬀerent br a nds like the levlor, Bali and vista. The sample b o oks ar e actual pieces of the blind so that th ey know ex a ctly what they a re getting. 5. Whic h are cr itica l factors that are imp ortant so as to b e successful in the selection pro cess? A: I ’d say all the factors of qualifying the customer would be the most signiﬁcant in the sele c tion pro cess b ecause if I know 34 how muc h the c us tomer is willing to sp end and if I know wha t he’d like the s a le would b e v ery s uccessful. 6. What really helps the custo mer in deciding from the diﬀerent types? A: Normally I’d give the customer the price and that r e ally would help them in deciding. Lik e fo r exa mple in Vijay a’s c a se since she wan ted arched vertical blinds. I kind of c ame to know that either she had lot money to sp end or that she did not know wha t she was up agains t. So when she c ame in I told her that it would cost her 800 dollar s a window and that made her decide that she did not wan t it So the price is one of the impo rtant factor s. W e really ca ll tha t brack eting the customer that is nothing but ge tting a feel of how muc h the customer is willing to sp end. 7. So if the type of pro duct is known and then how do you pro ceed? A: So of someone calls me and says that they’d like a r ched verticals I call up the vendor a nd ﬁnd out the price for standard window and g ive it to the customer so then the customer either gives his real meas ur ements or decides the window trea tment is to o co s tly . 8. Are blinds sp eciﬁc to a particular ro o m? A: Not really but when using a cer tain kind of blinds in particular it might not work out that go o d. Like if you put cellular in k itc hen then c leaning them mig ht be diﬃcult they ar e b etter in bedr o o m. When qualifying the customer I’d tell them they better not use these in the kitchen. So there are adv antages and disadv antages w hich I tell the custo mer during the selection pro cess. 9. How do you make sure the customer choice is corr ect? A: I tell them that sp ecial order ca nnot b e returne d that helps in dec iding for sure what they really w ant. 10. Ho w long do es it take to s e lect a sp ecial or der pro duct? A: F or sp ecial order pro duct it might take me a bo ut ha lf-hour to get a ll the infor mation that I need to k now and give them all the information that they need to know a nd co mplete the job. 11. When qualifying the customer what is the kind of informa tio n you are trying to get? A: The price, type o f pro duct, mea surements and the co lor. 12. If the c ustomer has a speciﬁc request what do you do? A: I’ll have to call the vendor and ask them if they can deliver the req ue s ted feature and they c an acco mmo da te it. 13. When I make a selection what do? Do you contact the vendor for av a il- ability? A: Y es after I get the selection then I contact the vendor to chec k if they hav e it in stock a nd ﬁnd how lo ng it will tak e for them to g et it. Normally it takes ab out 7 to ten weeks. 14. If the vendor ha s discontin ued the product? What is your next step? A: If I call up the vendor and ﬁnd that they have discontin ued the pro duct, 35 then may b e I’ll try and see if the other stores hav e that product if not then I just hav e to ask the customer to select another pr o duct. 15. What kind o f details do you give the vendor when you asking him for av ailability? A: I have to g ive details reg arding each blind. I have to g ive them the sku n umber for the pr o duct, the meas ur ements and the quantit y of the pro duct. 16. Ho w do you contact the vendor? If you a re not able to r each the vendor what do you do? A: I contact the vendor b y lo oking up the nu mber in the price b o ok. During norma l business hours I call the vendor and if I’m not able to contact I leav e a num b er to contact me bac k I also give the information they need. Then I ask the customer to either contact me or I contact them. 17. Does the vendor ha ve a n y delivery date sp eciﬁed? A: Y es, E ach vendor has a diﬀerent lead time spe c iﬁe s for each pr o duct. 18. Do you sugg est any accessor ies when the customer selects a blind? A: Everything that the customer needs for the blind c o mes co mplete in the order. No, I do not s ug gest any accesso ries to the customer . 3.4 KA Session # 4 3.4.1 Session Information Session #4: Designer generates q uotes fo r sp ecial order reques t KA Report Description: When customer selects a pr o duct from ca ta log and appr o aches the designer , the designer conﬁrms the selec tio n and genera tes q uo tes. Scenario Employ ed: Scenar io #2, Customer requests quotes on the pro duct selected from catalog. Do ma in Exp ert: J ulie, Senior Designer She is working for the Home Depo t Des ig n cen ter fo r the past 4 years. She helps the customer o f Home Dep ot to make the righ t pro duct selection and g enerates quotes for them to make a decision on the pro duct. Knowledge Engineer: Vija ya Balakrishna –In terview the designer Session Time and Date: 5.00 p.m. F eb 2 9th Session Location: Home Dep ot Design Center, Ro und Ro ck, Austin. 3.4.2 Session Purp ose T o obtain a detail knowledge ab out the work pro cess inv olved after the c ustomer selects a pro duct from catalog a nd appro aches the designer for quote, and what information the des igner needs in order to g enerate a quote. And to understand the r esp onsibilities a nd accomplishments of the designer to g enerate a quote. 36 3.4.3 Session Input T o fax the sample memo to domain exp ert, Julie s o tha t she is prepa r ed for the sch eduled session. Refer to ﬁgur e 2.11 for the memo sent to Julie on this session. Co nﬁrming the session with the p erformer Julie. Determining the KA techn ique to be used. Prepar ing the questions o n the sub ject ma tter to interview the do main exp ert. 3.4.4 Session Output T o understand the pro cess ﬂo w of how the designer understands the customer requirements in or de r to generate a quote for them. KA T echnique: W ork pro cess Analysis Type of Knowledge: Pr o cedural knowledge. This is used to o btain a detailed knowledge of the task in a step by step metho d a nd how each of these steps a re linked to acco mplish the tas k. Int e rview P lan: The total time 1hr a llo cated for the Interview is divide d betw een the ma in task s inv olved in this KA session. Refer to the ﬁg ure-2.10 below for planning map. 35 min 25 min Figure-2.1 0-Planning map for KA session#4 Memo sen t to Julie for KA sessio n#4 : T o: Ju lie, Senior Designer- H ome Dep ot F r om: Vijaya Ba lakrishna, Softwar e Engine er Hel lo, The Q& A session which I mentione d to e arlier may take up to 1 hr, we ar e sche dule d to me et you at 5.o o p.m on Thursday, the F eb 29th. I plan to c over two topics Cust omer appr o aches you after making a pr o duct sele ction Y ou gener ating quote for t he customer If ther e is any re fer enc e materials which I c an se e and use to understand the pr o c ess, it would b e very useful. Tha n ks a lot for c o-op er ating on this pr oje ct. Vijaya KA session questions : T a sk #1: Designer qualifying the customer 1. When a customer selects the blinds fr om catalog and a ppr oaches you for quote, wha t do you do next? A: I ask for the measurements o f t heir windows to g enerate a quote 2. What if the customer does not hav e measurements? A: I tell them that in o rder to generate a quote I need measure ments, but I give them ro ugh estimate on a n average window size, fr om that they can ﬁg ure o ut what price ra nge they ar e lo o king at. Whether it is for 1 window o r 10 windows. 37 But it is imp ortant for me to know the exac t measure ments to ge ne r ate a quote and actually we also go to the customers’ house and measure their windows 3. Quotes c an b e genera ted o n an av era g e size window, why do you ne e d measurements? A:If it is for the entire house, the window size v aries and of course the pr ice for the diﬀer e n t sizes and I ha ve to kno w that to giv e them a quo te. 4. What if the customer has measurements? A: I ask them if they ha ve measured their windows corre c tly , that is top, middle and b ottom to 1/ 8th of an inch and if so give them thr ee diﬀerent quotes. Refer to DOC -1 5. What if the customer ha s wrong mea surements? A: Most of the ca ses it is true, they often tend to measure length and width, but most of the windows are not p erfect ac r oss its length a nd width and so I sugg est them to re-measure . 6. Is the cor rect measurements r e quired in order to generate a quote? A: Actually not, r ough measurements for a ll the windows is eno ugh for me to generate a quote. 7. Do you tell the customer that Home Depot does measurements? A: Y es, I tell them that w e can measur e their house but there is a measur ement fee of $25 a nd if they decide to buy the pro duct then they get a credit for that a mount. 8. Once you ge t the window measurements, do you g o ahead and genera te a quote? A:I g ive them quotes on three diﬀerent pro ducts. 9. What if t he customer ha s made a wrong selection, do you let him/her be aw ar e ab out it? A: Y es, as a designer I hav e to unders tand what the customer really needs. I tell them the adv antages and disa dv a ntages of all the pro ducts and this is done b efore they make the decis io n on the pro duct. But a gain the customer would s e le ct a pro duct from the catalo g which is in trend, like select woo den blinds for all the windows in their house, and wo o den blind is not g o o d next to the kitchen sink and in the bathro om next to the tub. So I hav e to make them b e aw ar e a bo ut it a nd give them an alternative, the faux w o o d whic h is similar to the woo den blind a nd is also water res istant. 10. What kind of pro duct details that you conﬁrm with a customer b efore generating the quote? A: It dep ends. If the blinds selected are 2” hori- zontal, then I ment ion the co n trols that is asso ciated with this blind – tilt cord left, lift c ord r ight. And there is an option to reverse the controls. If the blinds is a 1” horiz o nt al, then there is wand tow a rds the left and lift tow ards the right. Again there is a pos sibilit y of r eversing the controls. Similarly , there ar e con tro l options fo r cellular and v erticals. I can note this down on a pap er for you. Refer to DOC 4 38 11. Is the pro ces s diﬀerent if the customer selects the wall pap er fro m the catalog and r equests a quo te? A: The pr o cess is muc h simpler, I just need to know how many double roles of wallpap er they requir e. But the rest of the pr o cess is same. T a sk #2: Designer gener ating quotes for the customer 12. Do you genera te a quote on the pr o duct selected ? A: Actually I g enerate quote on three diﬀerent pro ducts. 13. Wh y do you g ive the customer three diﬀerent quotes ? A: First of a ll I give them only three diﬀerent quotes b ecaus e they a re not over-whelmed to make a decisio n. And I may give them quotes on three diﬀerent pr o ducts so that they can diﬀerentiate b etw een the price r ange on the type of pro duct. I may g ive three diﬀ er ent quotes o n the same pro duct – but by three diﬀerent vendors. 14. On what three diﬀerent products do you generate the quotes ? A: It is normally the Rea l W o o d, Ce llula r, and the F aux W o o d for blinds. 15. Wh y is that these three pr o ducts a re chosen among others ? A: First of all these three pro ducts are the pr o ducts that are mostly chosen by customers. Th us they form a trend amo ng buyers. Next, these pro ducts oﬀer a g o o d mark-up v a lue to Home Depot. 16. Are there any speciﬁc situations where you generate more than three quotes ? A: Y es, b y this time I know if the customer is really serio us ab out making a purchase and if this also inv olves mo re pr o ﬁt fo r us, then I g enerate more than three quo tes. 17. Do all the v e ndo rs make the same kind o f pro duct ? A: Mos tly Y es , but there may be some vendors who has limited c ho ices in color s oﬀered b y vendors, design to ins tall the blind a nd the controls, a nd other things like 1. Limited choices in tap es for blinds 2 . Limited choices in blind-texture 3. Limited choices in installing the cords for the blinds 18. Where do you get the necessa ry information to gene r ate the quotes ? A: I consult the price b o ok to gener ate the quo te. The price b o ok lis ts all the vendors that we deal with and all the diﬀerent pr o ducts that they manuf acture, and als o the price rang e o n each of the pro duct on all window sizes. I use this information to gener ate a sp eciﬁc quo te. 19. Is your quote generation pro cess cur r ently a utomated ? A: No, I generate a quo te on a shee t of pap er and give it to the customer. 20. What kind o f information do es this quote contain ? A: I list all the window measurements (width * length) and the diﬀeren t vendors, Then I list the diﬀerent prices fo r each of these co mbinations. 39 21. Does the customer decide to buy the pro duct right aw ay? A: No, in most of the cases, they take this informatio n home to make a decision a nd then let us know if they ha ve decided to g o with us, or come bac k to us with more questions. They may a lso ask for a quote on a diﬀerent pro duct, by a diﬀerent vendor a lto gether. 22. If the customer decides to buy the pro duct, then wha t is the next step ? A: I ensure that they hav e the right measurements b efore I g enerate an inv oice. 3.5 KA Session # 5 3.5.1 Session Information Designer genera tes quotes for sp ecial or der request. When customer selects a pro duct from catalo g and appr oaches the designer , the designer conﬁr ms the selection and generates quotes. Scenario Employed: Scenario #2, Customer requests quotes on the pro duct sele c ted from cata lo g. Doma in Exp ert: Julie, Senior Designer She is w orking for the Home Dep ot Desig n center for the past 4 years. She helps the custo mer of Home Dep ot to mak e the r ight pro duct selection and generates quotes f o r them to make a decision on t he pro duct. Knowledge Engine e r: Vijay a Balakr ishna –Int e rview the designer . Sess ion Time and Date: 10.00 a .m. Mar ch 2nd Session Lo cation: Home Dep ot Design Center, Round Ro ck, Austin. 3.5.2 Session Purp ose T o understand the key strateg ies the desig ner uses to p erform th e tas k and generate task decomp osition a nd to ﬁnd out the r e sources inv olved in doing the task a nd sub tasks 3.5.3 Session Input 1. T o fax the sa mple memo to do ma in exp ert, Julie s o that she is prepar ed for the scheduled se ssion. Ref er to ﬁgure 2.13 for the memo sent to Julie on this sessio n 2. Conﬁrming the ses s ion with the p erfor mer Julie 3. Determining the K A technique to b e used. 4. Pr eparing the questions on the sub ject matter to in terv iew the do main exp ert. 40 3.5.4 Session Output T o Iden tify the task and s ubtasks inv olved in designer generating quo te for the customer. KA T echnique: T ask Analysis Type of Knowledge: Pr o cedural knowledge. This is used to o btain a detailed knowledge of the task in a step by step metho d a nd how each of these steps a re linked to acco mplish the tas k. Int e rview P lan: The total time 1hr a llo cated for the Interview is divide d betw een the ma in task s inv olved in this KA session. Refer to the ﬁg ure-2.12 below for planning map. Memo sent to Julie for KA sess io n#5: T o: J ulie, Senior Designer- Home Dep ot F r om: Vijaya Balakrish n a, Soft- war e Engine er, Hel lo, This fol low up Interview is to understand t he 1. T ask s and subtasks inv olved in qualifying the customer 2. T asks a nd subtasks in volved in g enerating the quote I think this In terview would take appr oximately 1 hr t o 1hr 15 minutes. We ha ve agr e e d to me et on Mar ch 2nd T uesday at 10. 00a.m. Thanks a lot for your time. Vijaya KA session questions : SubT a sk #1: Designer qualifying the c us tomer 1. What is your goa l whe n customer s approa c h you with their selections fo r a quote ? A: I hav e to qualify the customer and genera te a price quote which makes them buy the pr o duct. 2. What is the key accomplishment for you to p erfor m this job ? A: It is to qualify the cus to mer. That is to under stand the customer requirements so that I deliver what they rea lly wan t. 3. Do y ou have cus tomers approaching you for r equest all the time or is there a n y spe c iﬁc time of the day when they approach you? A: W e have customers walking int o the de s ign center all the time. But we are busy during the evenings and the weekends. 4. What is the prerequis ite for y ou to generate a quote ? A: I need the window meas urements, type of pro duct, and the color of the pro duct, to generate a quo te. 5. How mu ch time do es it take for you to generate a quo te? A: It may v ary from 5 minutes to 30 minutes. 41 6. Wh y do es the time v ary ? A: It dep ends, sometimes the customer may not be sure of the pro duct that they hav e chosen. It also dep ends on the nu m be r o f windows and the num b er of quo tes that I am gener ating. 7. Doe s y our task c hange if the customer has ma de a wrong selection ? A: Y es, I expla in to them the disadv antages of the pro duct that they hav e chosen a nd they hav e to star t the pro duct selec tio n pro cess all over again. 8. Do you need to be exp erie nce d in order to handle all the tasks after a pro duct is selected fro m the catalog? A: Selecting a pro duct fro m a catalog is a sp ecia l or der request and the desig ner has to kno w the deta ils of the ca talog pro ducts. If the designer is unable to ha ndle all the tasks, then he/she would co nsult the other ex per ienced des ig ners in fulﬁlling this s pecia l orde r reques t. 9. Doe s the num b er of years of ex per ience matter ? A: On a n average we need to hav e at lea st 6 months of exp erie nce. Every day we are asked new questions tha t we hav e never b een asked b efore. 10. Does your task change if the pro duct they have sele c ted inv olves one win- dow or ten windows ? A: No, it do es not change. But it takes more time to generate gener ate the ﬁgures. 11. Does y our task ch ange if the customer do es not hav e measurements ? A: Y es, I can give a roug h estimate on one w indow size. I ask them to co me back w ith precis e measure men ts in or der for me to gener ate a q uo te. 12. What is the av era ge window size ? A: It is 3 5’ * 64’ (width * height) 13. Ho w do you generate a price bas ed on this window size? A: W e have a price b o ok which we r efer to, in order to ﬁnd the price listings for a ll diﬀerent pro ducts, vendors and window sizes. 14. Ho w do you pr e sent this quote to the customer ? A: I write this informa- tion on a sheet of pa p er and g ive it to the customer . 15. If the customer has not meas ure their windo ws , what kind of assista nce do you o ﬀer to the cus tomer? A: W e tell them that we can measure the windows for them, but it w o uld cost them $25, whic h would b e adjusted tow ards their paymen t if they decided to buy the pro duct with us. Or I instruct them on how to mea sure the windows. 16. Ho w do you instruct the customer to measure their windows? A: I tell them to measure top, middle and b ottom to 1/8th of an inch – and as width * heig ht 17. If the custo mer has meas urements, doe s your task change? A: Y es , in this case, I just hav e to genera te a quote ba sed on the pro duct selec ted. I g en- erate quo tes for three diﬀerent pro ducts. If they are ﬁr m on one par ticular pro duct, I generate a quote on the same pro duct – but manu factured by three diﬀerent vendors. 42 18. Does your task change if the customer buys a pro duct to b e installed in a house, a partment or business? A: No, the pro cess follow ed is the same in all the thre e cases. 19. Are there an y situations where you approa ch y our sup ervisor in order to do this task? A: No, there ar e no such situations . 20. Are ther e a ny o ther per sons involv ed it ex e c uting this task? A: If the customer requests measurements to b e done by us, then I hav e to contact the mea surer to give the customer s informa tion. 21. Do you alwa y s hav e enoug h infor mation to execute the ta sk? A: No, not alwa ys. The customer will select a pro duct from a cata log and r equest for a specia l design. T o fulﬁll this request, I hav e to contact the v endor of this pr o duct to ﬁnd out if they carr y the desig n. 22. What if you a re not able to contact the vendor? A: I r ecord the customer information and later on co n tact the vendor. I then pass this information along to the customer at a later time. 23. Do you r equire customer informa tion in order to g enerate a quote? A: No, for gener ating a quote I do not need customer information. I nee d customer information o nly when decides to buy the pro duct. 24. Does the amount o f time sp en t on the customer dep end on the time of the day? A: I nor ma lly sp end more time at the end of the day b ecause I will be done with all the administr a tive task s. 3.6 KA Session # 6 3.6.1 Session Information Designer generates quo tes for spe cial or der request KA Rep ort Description: When customer selects a pro duct fro m catalog and approaches the designer, the designer conﬁrms the selectio n and gener ates quotes. Scenario E mployed: Scenar io #2 , Cus tomer requests q uo tes on the pro duct selected fro m catalog . Domain Exp ert: Julie, Senior Designer She is working for the Home Dep ot Desig n ce nter for the past 4 years. She helps the customer of Ho me Depo t to make the right pro duct selection and generates q uo tes for them to make a decision on the pro duct. Knowledge Engineer: Vijay a B a lakrishna –Interview the designe r (cont a ct Information) 43 Session Time and Date: 5.00 p.m. Mar ch 6th Session L o c ation: Home Dep ot Design Center, Round Ro ck, Austin. 3.6.2 Session Purp ose T o analyz e the task frequency , the task duratio n, the conditions to execute the task, the s equence of tasks, a nd to ﬁnd out the reso urces inv olved in doing the task a nd sub tasks 3.6.3 Session Input 1. T o fax the sa mple memo to domain ﬁgure 2.1 5 for the memo sen t to Julie on this se s sion exp ert, Julie so that she is pr epared for the scheduled session. Refer to 2. Conﬁrming the ses s ion with the p erfor mer Julie 3. Determining the K A technique to b e used. 4. Pr eparing the questions on the sub ject matter to in terv iew the do main exp ert. 3.6.4 Session Output T o ident ify the ter ms and resour ces used to do the ta sk, timing co nstraints in doing the task and the duration of each task. KA T echnique: T emp oral and Conceptual Analys is Type of K nowledge: Semantic knowledge. T o understa nd the ma jor con- cepts, deﬁnition and the voca bula ry used in the domain a nd the re la tionship betw een the tasks and the subtasks Int e rview P lan: The total time 1hr a llo cated for the Interview is divide d betw een temp oral analysis and the Conceptual analysis of scenar io#1. Refer to 2.14 b elow for planning on KA session 30 min 30 min Figure-2.1 4-Planning map for KA session#6 Memo sent to Julie for KA sess io n#6: Figure 2 .15: Memo for KA session#6 KA session questions : 44 T e mp or al Analysis 1. In how many cases do the custo mers come into the design center with the right s election? A: In most of the case s the cus to mers a r e not sure o f the selection until they buy the pro duct. 2. In how many instances do the customers come into the design center with bo th pro duct and also has the measur ement s? A: 8 0% of the time they do not hav e the measurements. 3. How muc h time do you sp end with a customer to make the right pr o duct selection? A: It v a ries to either 5 min utes to 30 min utes or it may ta ke few days. 4. What resour ces must b e in place before you ge nerate a quote? A: I need to hav e the measurements o f the windows, color of the product, type of pro duct and the window tr eatment selected. 5. What is the output of this task? A: Gener a ting three diﬀerent quotes fo r the cus to mer bas e d on the selections made. 6. How m uch time does it ta ke to genera te a quote? A: It may tak e from 5 min utes to 30 min utes, bas ed on the num b er of windo ws and diﬀeren t vendors a nd pr o ducts. 7. What is the frequency of g e nerating quote for a custom order pro duct? A: Ab out half the time we encounter custom order pr o duct selectio ns. 8. Is there any co n ting ency in doing your tas k? A: I canno t generate a quote if the customer do es no t have any measurements. Conceptual Analy sis: 1. What do y o u mean by qualifying a customer? A: T o understand the requirements based on who, what, when, wher e, and wh y . 2. How imp orta n t is this to generate a quote? A: On a scale of 1 to 10, it is a 10 b ecause this makes me realize what the customer really wan ts and how I can help the customer ma ke the right selection in order to buy the pro duct. 3. What is the term used for s e lecting a pro duct from a catalog? A: It is called a sp ecia l order pr o duct. 4. Will all the desig ners at the des ign center b e able to help the customer inv olving a sp ecial order request? A: The designers need to hav e atleast a minimum of 6 months exp erience in order to help the cus tomer with sp ecial order pro ducts . 45 5. Can the quo tes b e obtained ov er the telephone or internet? A: Currently we do not ha ve the pro cess of obtaining quo tes ov er the in ter net and w e do not encour a ge giving quotes ov er the tele pho ne. 6. What is the measure ment of the a verage window size? A: It is 35’ * 64 ’ (width * height). 7. How do you instruct the customer s to meas ure the windows? A: I instruct the customer to measure top, middle, bo ttom 1/8th of an inch and a lso give the customers, vendor catalo gs, which explains ho w to measure windo ws with pictures. Ther e are cata logs by diﬀ erent vendors like: Desig nView, Levolor. These catalogs hav e illustrations of the diﬀerent products that they manufacture and how to take measurements for their windows with pictures. 8. What do you re fer to in order to generate a quote for the customer? A: I refer the pr ice b o ok which lists all the vendors, the diﬀerent types o f pro ducts that they manuf a cture, a nd window sizes, and the price fo r ea ch window siz e . 9. Do all windo w sizes are entered in the price bo o k? A: The price for the window siz e s are rounded to the next highest even integer. 10. Ho w do you pres e n t the quo te to the customer ? A: I write the quote on a sheet of pap er and give it to the customer . Refer to DOC 2 11. What kind of information do es the quote contain? A: The q uote contains window mea surements for the diﬀerent pro ducts a nd their price range. 12. What are the compo nen ts of the ho rizontal blinds? A: The comp onents are: The headrail is used to hold the sla ts. The horizontal blinds can come with o r without v alance. And the v alance to cov er the headra il. 13. What a r e the comp onents used with V ertical B linds? A: The comp onents are: The headra il is used to hold the v anes. The vertical blinds can come with o r without v alance. And the v alance to cov er the headra il. 14. What ar e the controls av ailable with horizontal blinds? A: It dep ends. If it is a 2” blind, then the controls are tilt cord left, lift cord right. Or it may be the o ppo site. If it is a 1” horizontal, it has wand on the left, and the lift cord on the right. Again the o ppo site is p oss ible. If this is a cellular blind, there is a choice for the lift cord to b e either left or r ight. 15. What ar e the controls asso ciated with the vertical blinds? A: There is a choice to stack the blinds tow ards left, right or split it to either side. Controls are alwa ys o n the same side of the stac k. The wand is used to tilt as well as trav erse . 46 16. When do you realize tha t you need to g e nerate an inv oice? A: If I a m sure that customer is s ur e to buy the pro duct and has a ccurate measurements, than I genera te an in voice. 17. Wh y do you str ess o n customers getting a ccurate measur emen ts ? A: This concerns sp ecial o rder pro ducts, and the customers cannot return these sp ecial order pro ducts. The vendors will go ahead and deliv er the blinds with whatever measurements that the customer supplies. If the measurements are wrong, the vendors will not accept the pr o duct back. Therefore, w e stress that the customer measurements b e accura te. W e cast a shadow o f doubt to the customer by making them sig n a form which states that the measurements they ha ve taken are accurate. Only after that we or der the pro duct. Refer to DOC 6 18. Will the customers never b e able to r eturn the pro duct? A: Actually Home Dep ot has a po lic y to a ccept all returns . So we do accept return order pro ducts, but we do not let the customer know ab out it. This is b ecause we cannot return the blinds to the vendor and we hav e to mark down the pro duct, there by incurr ing a loss. 3.7 KA Session # 7 3.7.1 Session Information Session #7 : Designer genera ting in voice on the pro duct selected KA Rep ort Description: After the designer is sure that the customer ha s made the right selection and also has the rig h t measurements and is also r eady to buy the pro duct, than she would go a head and generate the inv oice Scenario E mploy ed: Scenario #3, Designer generates in voice base d on cus- tomer se lection. Domain Exp ert: Julie, Senior Designer She is working for the Home Dep ot Desig n ce nter for the past 4 years. She helps the customer of Ho me Depo t to make the right pro duct selection and generates q uo tes for them to make a decision on the pro duct. Knowledge E ngineer: Vija ya Balakr ishna –Interview the des igner Session Time and Date:10.00 a.m. Ma rch 1 4th Session L o c ation: Home Dep ot Design Center, Round Ro ck, Austin. 47 3.7.2 Session Purp ose T o ana lyze the pro c e s s the designer follows to do the tas ks, to deco mpo se these tasks and also to note the reso urces and the equipment used to accomplish this task. 3.7.3 Session Input 1. T o fax the sa mple memo to do ma in exp ert, Julie s o that she is prepar ed for the scheduled se ssion. Ref er to ﬁgure 2.17 for the memo sent to Julie on this sessio n 2. Conﬁrming the ses s ion with the p erfor mer Julie 3. Determining the K A technique to b e used. 4. Pr eparing the questions on the sub ject matter to in terv iew the do main exp ert. 3.7.4 Session Output T o ident ify the tasks and subtasks in volved when designer generates quote for the cus to mer KA T echnique: T ask Analysis Type of Knowledge: Pr o cedural knowledge. This is used to o btain a detailed knowledge of the task in a step by step metho d a nd how each of these steps a re linked to acco mplish the tas k. Int e rview Plan: The total time 1hr allo ca ted for the Interview is divided b e- t ween t wo to pics , one the designer gener ating the inv oice and the other cus tomer paying the bill. Memo sen t to Julie for KA sessio n#7 : T o: Ju lie, Senior Designer- H ome Dep ot F r om: Vijaya Ba lakrishna, Softwar e Engine er, Siemens Hel lo, As discusse d e arlier, In ou r next Interview I would like to c over t wo topics 1. T asks you do to gener ate an invoic e for the customer on a sp e cial or der pr o duct 2. and cus t omer p aying the bil l I think this Interview would take appr oximately 1 hr 15 min . We have agr e e d to me et on mar ch 14th T u esday at 10.00 a.m. Th anks again. Vijaya KA session questions : 1. When do you know you need to g enerate the in voice for the customer? A: When I am sure that customer has the right mea surements for a ll the 48 windows a nd has the right pro duct s election, tha n I gener ate the inv oic e for them. 2. How do y ou g e ne r ate the inv oice? A: I hav e an semi-a utomated system wherein I hav e to key in all the pr o duct se lections, the prices for each selection a nd the measur emen ts. 3. Do all designer ha ve access to the system? A: Y es, all designers hav e an access to the s ystem, If we hire new designer the computer guy ge ner ates an user-id for him/her 4. When do you input the customer information? A: This is one of the very ﬁrst tasks that I do when I am sure that the customer is going to buy the pro duct. I may do this even b efore I gener ate a q uote, dep ending o n how sure I am about the customer buying the pr o duct. But normally I do it after the cus tomer decides to buy the pro duct. 5. Can you explain in detail ab out the curr en t inv o ice g eneration system? A: I can use the sy stem by en ter ing my user –id and pas s word, next it brings up a screen wher e in I ca n lo ok up in to the existing orders or to create a new Inv oice, when I se le c t to create a new invoice, I need to enter the customer information, after I e n ter this informatio n, it brings up a inv oice form wher e I hav e to input measurements for each window, type of pr o duct, v endor info r mation, the co ntrols selected – three times f or every window one for the Receiv ing Dept, the installer, and the vendor. But if the customer has requested for mea surements I’ll hav e to type in t wice – o ne for the v endor a nd the other for the receiv ing department at Home Depot, so that the latter will hav e informa tion ab out the pr o duct that we or de r ed to the vendor. The invoice generated lo oks as if the information was automa ted, but I hav e ma n ually typed all the infor mation on every line. 6. How muc h time do es this take? A: This ta kes me at le a st one hour if I need to genera te an inv oice for the all th windows in the house. This pro cess is time consuming and I really wish if this could b e made muc h simpler. I always advise that the customer to come ba ck after a n hour to collect the invoice. 7. F r om wher e do you lo o k up the information in order to ent er it into the system? A: I ha ve the handwritten selections a nd measurements which I need e nsure that it is entered cor r ectly into the system. 8. Is there a po ssibility of en ter ing wrong da ta? A: Y es , there is alwa ys ro om for err or when entering thes e kind of details ab out the pro duct. This may lead to or dering the wrong selec tion. 9. Will it help if the s y stem gives you a choice to pick the pro duct, vendor, etc. when you generate an in voice A: Of course yes ! It will help us a nd sav e time. 49 10. Do you hav e any idea on how the curr en t system co uld improve? A: K ind oﬀ, I ca n write it for you. Refer to DOC 3 11. Can you expla in this? A: Y es, ﬁrs t of all I would like the scree n to display the av ailable vendors and than the type of the pr o ducts they make and the t yp e of fabric and I must be able to pick the window mea surements width * length instead of t y ping it. After this the system must give an option to generate quote or stage fo r r egister. If its only q uote, I don’t hav e to enter customer information but if it is s tage for register it must g ive an option to enter c us tomer information and than must give an option to select the color and controls for the t yp e of pr o duct and than the Information to where the customer wan ts the pr o duct to be deliv ere d and than to print the invoice. 12. After you e n ter all the pro duct information, what is the next thing you do? A: There is an option to gener a te quote, or stage for r egister. The stage for reg ister is used to print the in voice with the total pr ice on it. 13. After you g e nerate the inv oice, what ha pp ens? A: First of all the cus- tomer ha s to pay for it, next the information t yp ed to the vendor is faxed automatically b y the system and also to the Home dept receiving dept, and also to the measurer if he is g o ing to measure the windows. 14. What kind o f details go to the vendor? A: Most of the details such as the measurements, the controls and the type of pr o duct with the r e s pec tive SKU n umber 15. What is the SKU num b er? A: Ev er y type of pro duct such as ho r izontal, vinyl, cellular belo nging to a particular vendor ha s a SKU#. And SKU stands for Sto ck keeping Unit. 16. What details go to receiving dept? A: This is the v ague description of the pro duct. They are not concerned ab out the controls of the pro duct or the color, they just need to k now what pro duct, the quantit y a nd the vendor information, so if they re c eive the pro duct, they are sure that it is the o ne we ordered. 17. What information go es to the mea surer? A: All the informa tion we send to the vendor like the s iz e o f the window, the controls, the color, etc. 18. What do you do a fter you generate the inv o ice? A: yes I g ive it to him or her, there is a UPC co de prin ted on the in voice. I w alk with him to the cash register and the c ashier at the cash register scans the UP C co de and the cus to mer pays for it. 19. Do y ou do any veriﬁcation with the customer b efore they pay f o r t he pro duct? A: Y es I verify the scheduled date for de livery and tell them that it may v ar y a little a nd than again co nﬁrm the pr ice on the pr o duct befo re they pay for it. 50 20. If the customer has a discoun t coup on for a ho me dep o t pro duct, can b e taken care in the inv o ice? A: Y es I can do it or the cashier at the cash registry can also do it. In the system ther e is an option to mark down the pro duct, I have to e nter this discount informa tion in that place. 21. Do you give the discount on the total price? A: It is not p ossible, be c ause I can give disco un t only on the pro duct a nd no t on any serv ices like mea- surement and installation a nd the total amount includes all charges, so I lo ok up at the inv oice and mar k down, that is give a discount acc o rdingly . 22. Are there any par ticular paymen ts metho d that Home Dep ot follows? A: W e acce pt cash, cheque and all credit cards. 23. What are the critical facto r s for genera ting the quote? A: The custom- order measurements, o n the pro duct s elected – the amoun t of light that has to b e blo cked, how often the blinds are rais e d or lowered, and the color match. 24. If you ha ve to genera te an inv oice for custom-o rder wallpaper , is it the same pr o cess? A: Y es, it is m uch s impler b ecause I just hav e to enter a SKU n umber and the quan tity . Ther e are no other details inv olved in this case. It ca n v ary from 1 double-rolled wall pap er to any num b er of double-rolled wallpaper. 25. Once the bill is paid by the customer, do you still interact w ith the cus- tomer? A: It dep ends. If the cus tomer requests delivery to b e made to their house, and if they ar e satisﬁed with the pro duct, they will not call me, but the pro duct is delivered to Home Dep ot, than I will ca ll them once I get the pro duct. But mormally they call me if ther e is a pro blem with the pro duct. 26. If you need to access the customer informatio n, how do you do it? A: Our computer system provides an option when I lo gin, to see the existing orders, I lo o k up in to a par ticular o rder by the customer telephone num- ber . That brings up their inv o ic e, which I can use to access the necessa ry information. 3.8 KA Session # 8 3.8.1 Session Information Designer genera ting in voice on the pro duct selected. KA Rep ort Description: After the designer is sure that the customer has made the right s election and also has the right measurements a nd is also ready to buy the pr o duct, than she would go ahead and gene r ate the inv oice. This session is fo cussed on the 51 timing issues, the dur ation, the ter ms used and the pre-conditio ns and the p ost conditions to do the task. Scenario E mploy ed: Scenario #3, Designer generates in voice base d on cus- tomer se lection. Domain Exp ert: Julie, Senior Designer She is working for the Home Dep ot Design center for the pa st 4 years. She helps the customer of Home D e p ot to make the righ t pro duct selection and g enerates quotes for them to make a decision o n the pro duct. Knowledge Engineer: Vijay a B a lakrishna –Interview the designe r (cont a ct Information) Session Time and Date: 6.00 p.m. Mar ch 20 th Session L o c ation: Home Dep ot Design Center, Round Ro ck, Austin. 3.8.2 Session Purp ose T o ana lyze the pro c e s s the designer follows to do the tas ks, to deco mpo se these tasks and also to note the reso urces and the equipment used to accomplish this task. 3.8.3 Session Input 1. T o fax the sa mple memo to do ma in exp ert, Julie s o that she is prepar ed for the scheduled sess ion. Refer to ﬁgure 2.5.19 for the memo sent to Julie on this sessio n 2. Conﬁrming the ses s ion with the p erfor mer Julie 3. Determining the K A technique to b e used. 4. Pr eparing the questions on the sub ject matter to in terv iew the do main exp ert. 3.8.4 Session Output T o identify the terms and resources in the tasks and the pr e a nd po st conditions of the tasks and the frequency of e a ch task . KA T echnique: T emp oral and Conceptual Analys is Type of K nowledge: Semantic knowledge. T o understa nd the ma jor con- cepts, deﬁnition and the voca bula ry used in the domain a nd the re la tionship betw een the tasks and the subtasks 52 Int e rview Plan: The total time 1hr allo ca ted for the Interview is divided b e- t ween t wo to pics , one the designer gener ating the inv oice and the other cus tomer paying the bill. Refer to ﬁgure 2.18 for the KA planning o n ses s ion 8 Memo sent to Julie for KA sess io n#8: T o: J ulie, Senior Designer- Home Dep ot F r om: Vijaya Balakrish n a, Soft- war e Engine er, Siemens Hi Ju lie, I would b e c overing t he same t opic as discusse d last time, but n ow in this session I’l l b e c overing the details such as fr e quency, the dur ation to do the task and the t erms and r esour c es use d in doing this task. We have de cide d to me et on Ma rc h 20 t h at 6. o o p.m. Thanks agai n for your time. T hanks, Vijaya 3.8.5 T emp oral Analysis 1. How long do es it take you to g enerate an inv oice for blinds? A: based on the num b er of windows, I ma y ta ke from 15 minutes to 1 hr. this is bec ause I have to key in the s a me infor mation three times for e a ch window. 2. How long do es it take you to genera te an inv oice fo r wall pap er? A: it may ta ke up to 5 to 15 min utes b ecause no deta ils are inv olved 3. What is the preconditio n for you to genera te a Invoice for the blinds? A: The r ight meas urements a nd the r ight pr o duct selection. The c ustomer needs to b e sure o f his selections 4. What is the precondition for you to gener ate a Inv oice for the wall pap ers? A: The quantit y of double r oles they r equire a nd o f cour se the selected pro duct 5. What ar e the reso urces used to g enerate the inv oice? A: I use the computer system to type in the details of the pro duct selected, and to generate the Inv oice. I’ll use the price b o ok to verify the SKU# a nd the pric e for the particular measurement, a nd I’ll loo k in to the for m where I have customer measurements. 6. Do you use the sa me reso urces for generating in voice for the wallpap er? A: Mos tly yes, but I don’t need the measur ements, I use the price bo o k for SKU# and the pr ic e list, and in the same s creen in the system I’ll enter the qua nt ity and the SKU#. 7. What are the diﬀere nt tasks tha t ta ke place a fter you generate the inv o ice? A: I hand ov er the in voice to c ustomer to pay , a fter the custo mer pays the bill, a fax is automa tica lly faxed to the vendor by the sys tem a nd also to the r eceiving department a nd if the customer has requested for Installation, a fax is also s e n t to him a uto matically . 53 8. Are there an y problems y ou encoun ter while doing this ta sk? A: Y es sometimes, like in o ne case one of the designer g ave a wrong quote for the customer, that is actually for a lesse r price and once w e r e alized this we called the customer, but cus to mer wan ted to the pro duct on the earlier stated quote. 9. How do you handle these problems? A: In this case w e gav e a discount of 20% a nd did not charge the measur emen t fee. 10. Do you gener a te the inv oice with the lesser price? A: W e generate the inv oice on the actual price but we give a discount on certa in pro ducts which is reﬂec ted on the invoice. 11. What if the man ufacturer is not a ble to de liver the pr o duct on time? A: This happ ens some times, that is the r e a son we give the delivery date betw een 2 to 3 weeks. But if it extends even this time fr a me, than the vendor will notify me a nd I’ll ca ll up the customer and let them know. The time frame for delivery is not in m y hands, I alwa ys make this v er y clear to the customer and tell them a gain about it when they ar e ab out to pay the bill. 3.8.6 Conceptual Analysis 1. What do es SKU stand for ? A: It is the stock keeping unit, It can b e for all pr o ducts in the case of ins to ck pro duct o r it o n the type of pr o duct for a par ticular vendor for a sp ecia l or der pro duct 2. Do all the desig ner generate an inv oice for the s pecia l o rder pro duct? A: No they need to hav e at lea st 6 months of exp erie nce to do this on their own bec a use of the details inv olved, but they can alwa ys do with a ssistance from the mo r e exp erienced designer. 3. Are there an y s itua tions where is the les s exp erienced desig ner does not hav e any a ssistance from the mor e experienc e d designer? A: W e a lwa ys make sur e that there are at lea st tw o desig ne r s at all times, o ne to b e well exp erienced and the other pe r son may b e new in the store. 4. How do you sc hedule the delivery da te? A: if the customer has o nly requested the products but not installation, than we tell them 2 to 3 weeks, actually the system automatically gene r ates the scheduled date for delivery based on the vendors and the pro duct selected. But if the customer has reques ted for installation, than we give them a time fra me betw een 3 to 4 weeks. This is b ecaus e once the Installer gets the pro ducts, depe nding o n his schedule he’ll install the pro ducts. 5. What is the ro le of the Home dep ot receiving department? A: The re- ceiving department needs to know what pr o ducts hav e b een ordered fr om 54 all the departments a t Home Dep ot, b eca use if they receive the shipment they must b e in a p ositio n to verify it. 6. What are the details that you e nter in the sys tem? A: The v endor name (Like Gerb er, Bali, Levolor), t yp e o f pro duct that the v endor se le cted (like verticals, r eal w o o d 1”, cellular , faux woo d) and their respective SKU num b ers, type of fabric (like solitaire, midnight), measur emen ts of the windows, customer color and co ntrol choices, and then I do Stage- 4 register. 7. What is Stag e-4 regis ter ? A: This means I hav e e ntered all the selections and I am now g enerating the inv oice. This will also print a n inv oice rep ort which I give to the customer. 8. Do you have a sa mple inv oice that I can do cument? A: Y es , with customer per mission. Refer to DOC 5 3.9 KA Session # 9 3.9.1 Session Information KA Sessio n T itle: Discuss the pro cess at design center, to entertain customer ’s request for meas ur ement . Scenario Employ ed: Scena rio #4 Doma in Exp ert: Lisa Title: Systems Mana ger Employ er : Home Dep ot Ba ckground Informa tion: Lisa has be e n working a t Home Depot for the last seven years and is exp erienced in handling retail problems and managing the staﬀ working at the Design Center. She is manager ially resp onsible for all the tasks p erformed a t the Design Center. Lisa a lso decides on the p olicy and pro ductivity matters. Session Time a nd Date: 3/ 0 4/00 7:0 0 p.m. Session Lo cation: T raining ro o m at Home Dep ot, Round Ro ck 3.9.2 Session Purp ose T o ﬁnd o ut details of ac tiv ities p erfor med at the Design Center to entertain and pro cess the customer’s r equest to hav e the site mea sured by measurer of Home Depo t. 3.9.3 Session Input The Knowledge Enginee r schedules an app ointmen t with the domain exp ert and prepares for the ses s ion. He reads and tries to acquire more informatio n o n the techn iques that ca n b e used in this session. A list of questionnair e was prepared by the knowledge engineer and sent to the domain exper t so that the domain 55 exp ert is familiar with the pro cess and the type o f questions that will b e p ose d to them during this se s sion. Arrangements were made to audio tap e the s ession. 3.9.4 Session Output A detailed understanding of the work pro c e sses inv olved aro und pro c essing cus- tomer’s r equest for taking measur ement s of the site. T o outline any diﬀer e nc e s in the wa y selection of a pro duct is pro cessed if the customer comes without the measurements. 3.9.5 KA T ec hnique(s) W o r k Pr o cess Analysis, T ask Analysis, and T emp or al Analysis Type of Knowledge: P r o cedural knowledge Session Rep or t Date: 03 /04/0 0 5:00 p.m. Repo rt: 1. What do es the designer do w he n they come across a customer who do esn’t hav e mea surements? A: The designer w o uld either g ive a s a mple blind from the sto ck or ask the customer to g et the meas urements and come back a gain to the Desig n Cent er. If the customer has s ome idea ab out ho w to take measur emen ts the designer would hand ov e r a sa mple blind. In cas e the customer has no idea abo ut the siz e of windo w, the designer will teach the customer how to ta ke measurement s of the windows and ask for an accur acy of 1/8th inch. Its impor tant to hav e the exact mea s urements for the blinds befor e the designer cuts the blind. The desig ner w o uld a ls o let customer know ab out the meas ur ement and installa tion prog ram tha t the design cen ter oﬀers o n an extra charge. 2. What do you do if the custo mer do e sn’t wan t to go for measurements pro- gram? A: The designer takes the customer’s information like the name, address a nd the phone num be r . The des ig ner gives the customer a measur - men t form that ha s to be ﬁlled a nd sig ned by the customer. The customer comes bac k with the measurements and the desig ner pr o cesses the order further. The designer usually do esn’t qualify the customer if they don’t hav e a ccurate measurement. Opens a request that is faxed to the mea- surer. The customer go es to the front desk and pays for the serv ice. T he measurer co n ta cts the customer a nd ﬁxes an a ppo in tmen t and turns up at the site and takes the measurements of the necessa ry , windows etc. 3. What happens after the cus to mer dec ide s to go for the measuremen ts and installation progra m? A: After the customer decides for the service 56 the design center employ e e takes customer’s persona l informa tion in the system and the SK U o f the item they intend to buy . If the recor d of their information pre- exists in the system, the designer veriﬁes the informa tio n. The desig ner prints the information and a n inv oice is generated which is handed to the customer. The customer ta kes the inv o ice goes to the cashier at the front desk and pays. As so on as the customer’s payment is received a r equest is op ened for the customer and fax is sent to the measurer. The meas urer calls the customer within tw o business days and sets up a time that is conv enient for b oth. After the mea s urer mea sures, he faxes the measuremen ts back to the designer. The designer generates the quote for the pro duct a nd installa tion. The quote has all the details ab out the pro duct and the exa ct sizes. 4. Doe s Design Center hav e its own mea surer? A: Y es, the mea surer and the installer work only for the design centers of Home Depot. He covers all the Home Dep ots in Austin and has to entertain service r equests from all lo cations. 5. How do es the de s igner c o nv ey sp ecial requests of timing that comes from customer to the measur er? A: The designer enters the sp ecial timing request from the cus to mer in the system that is faxed to the measurer. F or instance, the customer co uld be out of town for t wo days etc. or the measurer might b e busy on that day . The desig ner do es not have a ny co nt rol ov er the mea surer’s schedule and it is b etw een the customer and the measurer to a gree on a suitable date and time. The measur er ha s an agr e emen t with the Home Dep ot to contact the customer within t wo days. 6. In case of custom or der, how do es the desig ner se nd the measurements to the vendor? A: The designer faxes the measurements along with the pro duct t yp e and details to the vendor a nd the pr o duct is delivered to the installer who g o es and installs the pro duct at the site. 7. What if customer would lik e to g o only for installation and not for measure- men ts or vice versa? A: The customer can c ho o se either for ‘mea surements and insta lla tion’ or ‘measurements only’. As a p olicy the insta lle r do esn’t install pr o ducts, whic h w a s not measured by him. This is becaus e he can’t guarantee the installa tion if the mea surements hav e not be en taken correctly . The home dep ot follows this p olicy . 8. Where ar e meas ur ement s recor de d after it is received from the mea surer? A: After desig ner r eceives the measurements form measur er it is keyreced int o the system. All the measurements along the customer information are recorded in the system. 57 3.10 KA Rep ort #10 3.10.1 Session Information Session Title: T o discuss the pro cess of installation, deliv ery and return of pro ducts. Scenario Employed: Scenarios #4 a nd #5 Domain E xpe r t: Lisa Ti- tle: Systems Manager Employ e r: Home Depot. Bac kg round Infor ma tion: Lisa has b een working at Home Depot for the last seven years and is ex per ienced in handling retail pro blems and mana ging the staﬀ working at the Design Cen ter. She is resp onsible for all the tasks p erformed at the Design Center. Lisa also decides o n the p olic y a nd pro ductivity matters, which concer ns customer ’s satis- faction. Sess ion Time and Date: 3 /18/0 0 5:00 p.m. Sess io n Lo cation: T ra ining ro om at Home Dep ot, Ro und Ro ck 3.10.2 Session Purp ose T o ﬁnd out the de ta ils abo ut th e w ay customer’s request for installation is pro cessed. 3.10.3 Session Input The KE schedules an a ppo in tment with the domain ex per t and prepar e s for the session. He r eads a nd tries to acquire mo re informa tion o n the techniques that can b e used in this ses sion. A list of questionnair e w as pre pa red and sent to the domain ex p er t to keep the domain exp ert familiar with the pro ces s and the t yp e o f questions whic h will be p osed to them during this session. 3.10.4 Session Output A detailed understanding of the work pro c e sses inv olved aro und pro c essing cus- tomer’s r equest for taking measur ement s of the site. T o outline any diﬀer e nc e s in the wa y selection of a pro duct is pro cessed if the customer comes without the measurements. KA T echnique(s): W or k Pro cess Analysis , T ask Analysis, a nd T empo ral Analysis Type of Knowledge: P r o cedural knowledge Session Rep or t Date: 03 /19/0 0 Repo rt: 1. How do es the customer co me to know ab out the installation progr a m? 58 A: There a re banners in the design center tha t inform the customer ab out it. It says Pr ofessional installation a v ail able . F or the details of the progra m the cus tomer will ha ve to approach the designer. The designer explains ab out the pr ogram to the custo mer and lets him know the rates charged by the desig n center. 2. How is the informa tion ab out the delivery of the pr o duct pass ed to the delivery team? A: F or a custom order the options ar e indicated to the vendor. The cus- tomer has thr e e delivery o ptions of e ither a ccepting the delivery at home or to the insta lle r. The pro duct c a n also b e delivered at the Home Dep ot. The front desk usually handles this. So metimes they buy other items from Home Depot to o a nd a s k for combined delivery . The delivery is made by a sp ecial department which ta kes care of all this, I am not sure a b out the detailed w orking o f this department. 3. What happ ens after the v endor has delivered the pro duct. (In case of custom o rder). A: The ins taller installs the pro duct only if the measure- men ts were taken by Home Dep ot. This is b ecause the installer can’t guarantee the installation if the measurements w ere not taken correctly . The measurer appro aches the customer to ﬁx up a s uitable time and go es and installs the pro duct. 4. What is do ne to pro duct r eturned by the customer? A: The c us tomer returns the pr o duct and receives the money back at the se rvice desk. At the end o f each day the des igner of the desig n center picks up the r eturned pro duct from the front des k. The designer insp ects eac h product. If the pro duct is an in sto ck pro duct a nd if it s condition is go o d, it’s plac e d back on the shelf. If the pro duct is an ex-sto ck pr o duct, an R TV (Return to V endor) is generated. Then the designe r marks down each pro duct. The returned items ar e mark e d down t o zero as o ppo sed to a certain per centage of the v alue in c a se of discounts. The insto ck items which a re not in co nditio n to b e pla ced back on shelf is trashed. 1 5. What do you mean by mar king down of the pro duct? A: Mar k do wn means lowering the v alue o f the product. Mar kdowns are done in t wo case s . First, to give the c ustomer some disco un t at the time of purchase. This is nothing but disco un t. Second, to account for the lost, damaged or retur ned pr o duct. In the second ca se, the g o o ds are marked down to zero, meaning the pro duct has b een lost completely in v a lue . There is a reg ister wherein the details of the pro duct SK U/UPC Num b er and item descr iptions ar e recorded. The reaso n for damage is also recorded. (A sample form is attached with the rep ort). 2 1 The return of pro duct is not par t of the domain and therefore not mo deled. Refer to the domain men tal m odel. 2 Same as abov e. 59 6. Can customer pay for the pr o duct b efore the measurements a re taken, to av oid coming aga in to Home Dep ot? A: Y es. The customer can pay by his c redit car d in that ca se b ecause the exact measurements a re no t tak en and there fore the exact amount can’t be calculated. This ha ppens quite often, the customer gives his credit card num b er and the expira tion date o f the ca rd to the designer and after the designer g e ts the measur emen ts from the ins ta ller he/she charges the credit card of the customer. 7. Doe s Ho me Dep ot deliver in sto ck pro duct? A: Y es. Home Dep ot delivers the pro duct to cus to mers. The customer go es to the cashier and pays for the delivery along with the pa y men t for the product and the g o o ds are tak en out in the trucks to the customer’s lo cation. 8. How do es the delivery work in case of custom or der? A: In ca se o f sp ecial order, a t the end of o rder the designer has to select the destination of delivery . There ar e three options of shipping the pro duct delivered. The des tination could either be Home Dep ot or Installer or to the customer . By this time designer kno ws where the pro duct has to b e shipp ed a nd picks the correct resp onse. The customer could hav e either selec ted o f a self-installation or for ins tallation by the Ho me Dep ot installer. If the custo mer re quested for delivery to their home address the designer takes do w n the instr uctions from the customer. This includes the day and time they wan t the items to b e delivered. The designer a ls o a s ks if the cus tomer is comfor table with the items be ing left a t the do ors tep in case they a re not presen t at the time o f delivery . In case they are no t, the go o ds are delivered to Home Dep ot and the customer comes and picks up the pro duct fro m Home Dep ot. The Home Dep ot staﬀ infor ms the customer when the items arr ive. As the third option, the items ca n b e delivered to the installer and the insta lle r schedules with the customer fo r installation. 9. How doe s Home Dep ot track the order s de livered to Home Dep ot for its customers? And how do they inform the customer? A: The Inv entory depar tmen t receives the o rder and informs the service desk and design c en ter ab out the receipt. Service desk calls the customer and informs them that the or der has b een received. The custo mer c o mes to the service desk with the receipt of paymen t made to pick the items. The s ystem is up dated when the customer picks up the order. 10. Does the pro cess of selection change if the customer do esn’t hav e mea- surements? A: The custo mer is adv ised by the des igner to get the measurements b e- fore des ig ner can help them. The designer tells the customer how to take 60 measurements so that they can ge t the rig ht measur ement s . Often cus- tomers re turn with the measurements. The mea surements are imp orta nt in product selection. The designer gets an idea what the customer’s site is like and is able to he lp s ell an appropr iate pro duct to the custo mers needs. This is one of the wa ys to qua lify the customer . 11. What are the c ha nges that can be made in the system to improv e it further? A: Every inv oice ha s an estimated date o f ar riv al at the end o f it. The estimated arriv al date could v ary dep ending on the t yp e of pro duct and the sp ecialty requirements, which go es along with it. F or e.g. the ar ches are takes time to cut, pack and deliver. The es tima ted date of arr iv a l often confuses the customer and wro ngly sets the cus to mers exp ectatio n. Another pr oblem that o ccurs is the break down of communication channel betw een the design center a nd the v endo r. The fax is either los t on the wa y or the receptions are not clear . The designer doesn’t come to know ab out the lo ss of infor ma tion during the c omm unication. It will b e go o d to hav e a b etter co mm unica tion wher e the desig n center employees can chec k the status of a ny o rder. The designers w o uld like to know if the customer r eceived the pro duct and also ﬁnd out if the vendor had received the comm unication and is proc essing the order at their end. The design center employ ees would like to keep track of ce r tain o rders and the system should b e able to pr ovide them with some kind o f automation to achieve this, so that the designer can act whenev er there is a comm unication breakdown. 61 Chapter 4 Kno wledge Mo dels 4.1 Ov erview The knowledge mo dels are representations of the information obtained in the KA sessions r egarding the pro blem domain. An atom o f information in the do main would fo cus on a concept in the doma in, it could b e a task, resource, state, even t, or a r elationship. The deﬁnitio n of a n atom would include the na me, synonyms, descr iption etc similar to a do main dictio nary . These atoms can be tasks, resour ces, sta tes , even ts, or relatio nship. A co llection on the other hand would provide the structur e and relationships b etw een a toms of information. Examples of co llections ar e task templates, task hier a rchies, sequence diagrams , state c ha rts, etc. F ollowing are the diﬀere nt kinds o f models that we will b e deriving from our KA se ssions. Her e is the br ie f descr iption of these mo dels. 1. Ev ent T race: An e ven t trace will repres ent the interaction b etw een re- sources alo ng a timeline this interaction may b e with regar ds to da ta , even ts o r task s . 2. T ask Decomp os ition: This model represent s the task bre a kdown in the work pr o cess a nd identiﬁes a ll the subta sks for each sup er task. This diagram represe nts all the tasks that take place in the problem do main. It also shows the r elation b etw een tw o tasks, lik e what a re a task’s sup e r task o r its sub ta sks. 3. T ask Hierarc hies : The task hierar ch y dia grams will sp ecify the diﬀerent wa ys to p erform work pr o cess o r tasks. 4. T ask T e mplates: The task templates will ha ve all the infor mation re- garding a task. It consists of details such as the frequency , duration, resource s, input-output, pre and po st co nditions and its relatio n to other tasks. 62 5. Concept Hierarc h y: The co ncept hierar ch y mo dels r epresent the ’kind of ’ relatio ns b etw een the res ources. 6. Concept Maps: These models are also known as the entit y relation diagrams. These models ar e us e d to characterize attributes of reso urces and relationships b etw een data elements. 7. Domain Dictio nary: The doma in dictiona ry is a deﬁnition o f a co ncept in the domain. 8. Do cument Analysis : The do c umen t analysis has the do cument re- sources that were used for mo de ling . 63 4.2 Mo d els 4.2.1 Mo del: Designer CA 01 Knowledge Mo del Num be r : Designer CA 01 Mo del Des c r iption: Conceptual Hiera rch y of Resources KA Source: K A Session Repo rt #1 Mo del Persp ective: Designer Date: 04/03/ 00 Figure 4 .1: Conc e ptual Hierar ch y o f Resour ces 64 4.2.2 Mo del: Designer CA 33 Knowledge Mo del Num be r : Designer CA 33 Mo del Des c r iption: Concept Map KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Figure 4 .2: Conc e pt Map 65 4.2.3 Mo del: Designer CA 08 Knowledge Mo del Num be r : Designer CA 09 Mo del Description: Entit y r elationship b etw een data concepts of the Quote Sheet KA Source: K A Session #6 Mo del Persp ective: Designer Date: 04/03/ 00 Figure 4 .3: Entit y relations hip b et ween data conce pts of the Quote Sheet 66 4.2.3. 1 Mo del : Designer CA 09 Knowledge Mo del Num be r : Designer CA 09 Mo del Description: Entit y r elationship b etw een data concepts of the Quote Sheet KA Source: K A Session #6 Mo del Persp ective: Designer Date: 04/03/ 00 Figure 4 .4: Entit y relations hip b et ween data conce pts of the Quote Sheet 67 4.2.4 Mo del: Designer CA 10 Knowledge Mo del Num be r : Designer CA 10 Mo del Description: En tity relationship betw een data concepts of Accurate Measurement for m KA Source: K A Session #6 Mo del Persp ective: Designer Date: 04/03/ 00 Figure 4.5: Entit y r elationship b et ween data concepts o f Accur ate Measurement form 68 4.2.5 Mo del: Designer CA 18 Knowledge Mo del Num be r : Designer CA 18 Mo del Des c r iption: Entit y rela tionship b et ween data co ncepts of Inv oice KA Source: K A Session #8 Mo del Persp ective: Designer Date: 04/03/ 00 Figure 4.6: Entit y r elationship b et ween data concepts o f Accur ate Measurement form 69 4.2.6 Mo del: Designer CA 19 Knowledge Mo del Num be r : Designer CA 19 Mo del Des c r iption: Concept hiera rch y for pe rformer Role KA Source: Intro duction, KA Sessio n #8 Mo del Persp ective: Designer Date: 04/03/ 00 Figure 4 .7: Concept hierarch y for p erformer Role 70 4.2.7 Mo del: Designer CA 22 Knowledge Mo del Num be r : Designer CA 22 Mo del Des c r iption: Concept Map representing p erformer roles. KA Source: K A Session Repo rt #9 Section 2.5.9 Mo del Persp ective: Designer Date: 03/04/ 00 Figure 4 .8: Concept Map representing p erformer roles 71 4.2.8 Mo del: Designer CA 24 Knowledge Mo del Num be r : Designer CA 24 Mo del Descriptio n: Concept Map representing p erfor mer r oles and infor ma- tion/entit y ﬂow KA Source: K A Session #10 Sectio n 2 .5.10 Mo del Persp ective: Designer Date: 04/09/ 00 Figure 4 .9: Concept Map representing p erformer role s and information/ en tit y ﬂow 72 4.2.9 Mo del: Designer CA 25 Mo del Description: Entit y relations hip diagra m KA Source: K A Session #10 Sectio n 2 .5.10 Mo del Persp ective: Designer Date: 04/09/ 00 Figure 4.10: Concept Map repr e sent ing p erformer roles and informa tio n/entit y ﬂow 73 4.2.10 Mo del: Designer T A 02 Knowledge Mo del Num be r : Designer T A 02 Mo del Des c r iption: T a sk Decomp osition for In sto ck Pro duct KA Source: K A Session Repo rt #1 Mo del Persp ective: Designer Date: 04/01/ 00 Figure 4 .11: T as k Decomp osition for In sto ck Pro duct 74 4.2.11 Mo del: Designer T A 03 Knowledge Mo del Num be r : Designer T A 03 Mo del Des c r iption: T a sk Hierarch y for: T a sk 2.1 - Check Inv entory T a s k 2 .2 - Contact V endo r T as k 3.1 - Pa y cashier KA Source: K A Session Repo rt #1 Mo del Persp ective: Designer Figure 4 .12: T ask Hierar ch y 75 4.2.12 Mo del: Designer T A 04 Knowledge Mo del Num be r : Designer T A 04 Mo del Des c r iption: T a sk template for task # 1.1.1 from Designer T A 02. KA Source: K A Session Repo rt #1 & KA Session Rep or t #2 Mo del Persp ective: Designer Date: 04/01/ 00 76 4.2.13 Mo del: Designer T A 05 Knowledge Mo del Num be r : Designer T A 05 Mo del Des c r iption: T a sk template for task # 1.1.2 from Designer T A 02. KA Source: K A Session Repo rt #1 & KA Session Rep or t #2 Mo del Persp ective: Designer Date: 04/01/ 00 T a sk Name: Obtain measurements and pr ice ra ng e T a sk Descr iption: Execution Duration: 1-5 minutes Execution F req uency: Discrete Pre-co ndition & Criticality Constra int pairs : Homeowner — High Pre-co ndition Statement: The p erson coming in to buy the product sho uld be the none other than the homeowner b eca use only the homeowner can make the dec is ion. Also the customer should get the meas urements then only the designer can help him. Having the price rang e what the customer plans to stay also will b e helpful. Post-condition Sta tement & Criticality Constra int Pairs: Infor mation Gained – High Post-condition Statement: When the designer knows the price r ange the se- lection pr o cess b eco mes faster having infor ma tion reg arding what the customer wan ts then he c a n assis t the customer w ith cor rect choices. Input Input Name: measur ement s and price r ange o btained Input Type: Event Criticality: High Sender of Input: Designer Output Output Name: Mea s urements Output Type : Data Criticality: High Sender of Output: Customer Output 77 Output Name: Pr ice ra nge Output Type : Data Criticality: Medium Sender of Output: Customer Lo cation T a s k is p erfor med: This task may b e per formed on the phone or at the Design Center Resources: Personnel: An y Desig n Center employ ee Equipment: None Sub T ask s: None 78 4.2.14 Mo del: Designer T A 18 Mo del Description: Scenar io Elicitation outcome ma trix KA Source: K A Session #4, section 2.5 .4 Mo del P resp ective: Designer Date: 04/06/ 00 Date Rev ised: W o r k Pro cess Performer Res o urces Used Int e ractions Result Problems E xe- cuting Qualify Cus - tomer to generate a Quote on the B linds or w a llpap er selected Designer Price b o ok No Qualiﬁes the customer by verifying 1. .Customer has made the right pro duct selection 2. .Customer has meas ured their windows and ha ve the right measure- men ts 1. .Customer do es not hav e measurements 2. .Customer not selected the r ight pro d- uct 3. .Customer do es not hav e the right measurements Qualifying Customer – Helps the customer t o make the righ t Blind or wall- pap er selection by giving all the adv a n- tages a nd details based on customer needs Designer Catalogs Y es- sometimes exp ert designer Customer requests a quote fro m the designer Generate rough esti- mation on an av erag e window siz e Designer Price b o ok No Helps the Customer on the price rang e of the pro duct they have selected 79 Qualifying the customer – Requests the customer for Windows mea- surements and provide deta ils for meas ur- ing windo ws and als o give an option of Home Depot measuring customer’s window Designer Nil Y es- If the customer requests Home De- po t f or measuring their win- dows, than the De- signer co- ordinates with In- staller 1. .E xpec ts the customer to r eturn with measurements 2. .Co- ordinates with the Installer for measur ing the customer windows V er iﬁes the customer has right measure- men ts Designer Nil No Decides to Generate Quotes for the pro duct selected Pro duct se- lected is discontin ued Generating quotes for the customer on a sheet of pap er Designer Price b o ok Y es- sometimes the expert designer on sp e- cial order pro ducts Generates quotes on three diﬀer en t pro ducts for the cus to mer As the pro cess is manual, it takes a lo ng time to gener- ate the quote 80 4.2.15 Mo del: Designer T A 20 Mo del Description: T ask Decomp osition KA Source: K A Session Repo rt#4 & K A Session Repor t #5(2.5.4 &2.5 .5) Mo del Persp ective: Designer Date: 04/06/ 00 Figure 4 .13: T as k Decomp osition 81 4.2.16 Mo del: Designer T A 26 Mo del Descr iption: T ask T emplate for T a sk 3.1 of Knowledge mo del n umber Designer T A 20 KA Source: K A Session Repo rt#5 (Section 2.5 .5) Mo del Persp ective: Designer Date: 04/06/ 00 T a sk Name: Co- o rdinate with the vendor on pro duct av ailability T a sk Description: If the customer is interested to buy the sp ecial order pro duct and requests a quote fro m the designer, the designer c he cks with the vendor on pro duct av ailability Execution Duration: 5 minutes Execution F req uency: Discrete Pre-co ndition & Criticality Constraint pairs: Customer requests quo te- Cus- tomer would like to buy the pro duct-High Pre-co ndition Statement: Customer r equests a quote on sp ecia l o rder pro d- uct. And is interested to but the pro duct Post-condition Statemen t & Criticality Constr aint Pairs: - Designer chec ks for the pro duct av ailability - High Post-condition Statement: The designer veriﬁes with the vendor if the pro d- uct selected b y the custo mer is av a ilable fo r delivery . Ba sed o n that infor mation, the des ig ner genera tes quote Input Input Name: Designer chec ks for pro duct av ailability , Selected pro ducts Input Type: data Criticality: Medium Sender of Input: desig ner veriﬁes with the vendor Output Output Na me: Pro duct av aila ble for delivery in the lead time, could not contact the vendor, pro duct not in sto ck or discontin ued Output Type : E vent Criticality: Medium Sender of Output: Designe r passe s the infor mation to customer 82 Lo cation T ask is per formed: This task is p erformed at the Design Center, Home Dep ot. Resources P ersonnel Designer , V e ndor Equipment Price b o ok 83 4.2.17 Mo del: Designer T A 45 Mo del Description: T ask template for tas k #3.0 KA Source: K A Session Repo rt #9 Section 2.5.9 Mo del Persp ective: Designer Date: 03/04/ 00 T a sk Name: Send Request to Mea surer T a sk Descr iption: Execution Duration: 1-10 minutes Execution F req uency: Discrete Pre-co ndition & Criticality Constra int pairs : T ask #2 Pre-co ndition Statement: The custo mer should have paid at the payment center. Post-condition Sta tement & Critica lity Constraint Pairs: The designer trans- mits the reque s t to the measur er. Post-condition Statement: The measur er receiv es the reques t along wit h customer information. Input Name: Customer Information Input Type: Entit y Criticality: High Sender of Input: Designer Output Name: Reques t T rans mitted Output Type : E vent Criticality: High Sender of Output: Designe r Lo cation T as k is per formed: The task is per formed a t the Design Cen ter. Resources Personnel: Designer Equipment: Communication System 1 1 Commun ication System: A r esource in the domain. Refer to domain dictionary f or deﬁnition 84 4.2.18 Mo del: Designer DOC 01 Mo del Description: Do cument F ramework F orm - Instr uctions to measure win- dows KA Source: K A sessio n #4, Sectio n 2.5 .4, At tachmen t DOC#1 Mo del Persp ective: Designer Date: 04/06/ 00 T o pic: Instructions to measure windows Key Co ncepts: 1. T o mea sure the window (width*Leng th) to the nearest 1/ 8th of an inc h. 2. T o mea sure top, middle a nd b ottom of the window SubT opic#1: Instructions for horizon tal or cellular blinds Key Co ncepts for SubT opic#1: 1. T o mea sure the smallest width and the Longest le ngth 2. T o mea sure windows with o r without head rail SubT opic#2: Instructions for V ertical Blinds Key Co ncepts for SubT opic#2: 1. T o mea sure the top w idth and the sma llest length of their window 2. T o mea sure windows with o r without the head r ail Reference: Do c#1 Do c #1 User : Custo mer uses the ab ove infor mation provided by the desig ner to measur e their windows 85 4.2.19 Mo del: Designer DOC 02 Knowledge Mo del Num be r : Designer DOC 02 Mo del Description: Document F ra mework F or m -Controls av a ilable with window Blinds KA Source: K A sessio n #4, Sectio n 2.5 .4, At tachmen t Doc #2 Mo del Persp ective: Designer Date: 04/06/ 00 T o pic: The diﬀerent controls av ailable with Window Blinds Key Co ncepts: 1. Standard co nt r ols 2. Reverse Standard controls SubT o pic#1: 2” Hor izontal blinds Key Concepts: Tilt cord and Lift cor d on rig ht side or left side of the window SubT o pic#2: 1” Hor izontal blinds Key conc e pts: Tilt wand and Lift w a nd on r ight side or le ft side of the window SubT o pic#3: Cellular blinds Key Co ncepts: Lift cor d on right o r left side of the window SubT o pic#4: V ertica l Blinds Key concepts: W ands on same side a s stack or on either side of the window Reference: Do c#2 Do c #2 User : Customer uses this infor mation provided by the desig ner to decide on what controls they need for their windows 86 4.2.20 Mo del: Designer CA 02 Knowledge Mo del Num be r : Designer CA 02 Mo del Des c r iption: Domain Dictionar y KA Source: K A Session Repo rt #1 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: W allpap er b o oks Related Knowledge Mo dels: Desig ner CA 01 Domain T er m Deﬁnition: The wallpaper bo o ks contain sample wallpaper and man ufacture r infor mation. The customer loo ks a t the b o ok to decide the t yp e o f pro duct he needs. Sample Usage: The wallpaper b o oks are not so rted by color but you will ﬁnd the colo r you w a n t in every b o ok . The ﬁrst thing the custo mer needs to decide on is the style of wallpaper that a r e con temp or a ry , tra ditional, kitc hen and bath. 87 4.2.21 Mo del: Designer CA 03 Knowledge Mo del Num be r : Designer CA 03 Mo del Des c r iption: Domain Dictionar y KA Source: K A Session Repo rt #1 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Mini Blind- Cutter Related Knowledge Mo dels: None Domain T erm Deﬁnition: The mini blind cutter is used at the design center to cut in sto ck blinds acco rding to c us tomer r e q uirement Sample Usage: Jane w a nt ed to buy in sto ck blinds but the size she wanted was not in sto ck. So Jimm y suggested that she could pic k a la rger size in the brand s he liked a nd that he could cut it to size she needed using the mini blind cutter. 88 4.2.22 Mo del: Designer CA 04 Knowledge Mo del Num be r : Designer CA 04 Mo del Des c r iption: Domain Dictionar y KA Source: K A Session Repo rt #2 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Price B o o k Related Knowledge Mo dels: None Domain T erm Deﬁnition: The price b o ok is the bo ok with all the pro duct sku’s the price and the vendor contact information. Sample Usa ge: When Jane w anted the Bali 1 –inch blinds in 40 by 6 0 and the size was not in sto ck, Jimmy lo oked at the Price b o ok and found the vendor nu m be r and called him up to a s k for the delivery date for that par ticular type. 89 4.2.23 Mo del: Designer CA 05 Knowledge Mo del Num be r : Designer CA 05 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #3 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Catalog Related Knowledge Mo dels: None Domain T erm Deﬁnition: The c atalogs are the product manuf acturer pub- lished magazines that hav e all the diﬀeren t mo dels man ufacture d by the com- pany for the customer to select. Sample Usage: When Cynthia wanted to decide on the diﬀere nt t yp es of the blinds then J ulie directed her tow ar ds the cata logs to see the diﬀerent t yp es catalogs stacked up on the shelf. 90 4.2.24 Mo del: Designer CA 06 Knowledge Mo del Num be r : Designer CA 06 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Mobile c a rt Context: Introduction Domain T erm Deﬁnition: The mobile cart is a like a wireless computer on the car t which is used by the desig n center employ ees to place order for inv entory and chec k inv entory etc. Sample Usage: When Jane want e d to buy a bali I-inch blinds in a quantit y of 1 0 then jimm y lo gged to the mobile ca rt and saw that they ha d only 8 in sto ck. 91 4.2.25 Mo del: Designer CA 26 Knowledge Mo del Num be r : Designer CA 26 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: Measurements Context: Select Pro duct Domain T erm Deﬁnition: Each blind ha s measurements, which are its length and the breadth they ar e in inc hes. These mea surements are used determine ﬁts customer’s window. So when a customer wan ts to buy the pro duct he needs to get the window s ize. Sample Usa ge: Ja ne measured her window fo und that it was 30 X 60 so with these measur ement s she wen t to the design ce n ter to reinstall her blind. 92 4.2.26 Mo del: Designer CA 27 Knowledge Mo del Num be r : Designer CA 27 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: V erticals Context: Select Pro duct Domain T erm Deﬁnition: This is a type of blind that is sold a t the design center. They hav e the individual pieces of the blinds arranged in a vertical pattern. Sample Usage: Ja ne lo o ked at the large window facing the sea and decided that vertical blinds there w o uld give her easy a ccess to the patio. 93 4.2.27 Mo del: Designer CA 28 Knowledge Mo del Num be r : Designer CA 28 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: Horizontal Context: Select Pro duct Domain T erm Deﬁnition: This is a type of blind that is sold a t the design center. They have the individual pieces o f the blinds arra nged in a horizontal pattern. Sample Usage: Jane lo oked at her kitchen window and felt that horizontal blinds ther e would b e e a sy to maintain. 94 4.2.28 Mo del: Designer CA 29 Knowledge Mo del Num be r : Designer CA 29 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: Cellular Context: Select Pro duct Domain T erm Deﬁnition: This is a type of blind that is sold a t the design center. They have the individual pieces o f the blinds arra nged in a horizontal pattern but they give more pr iv a c y than the horiz ontal. Sample Usa ge: J ane lo oked at her b edro om window and felt that cellular blinds ther e would give her eno ugh priv acy . 95 4.2.29 Mo del: Designer CA 30 Knowledge Mo del Num be r : Designer CA 30 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: Embossed Context: Select Pro duct Domain T er m Deﬁnition: This is a type of wallpap er that is s old at t he design center. They hav e the puﬀy impress io ns. Sample Usage: Jane felt that em b ossed wallpaper would g ive her mor e e le - gance to her dining r o om 96 4.2.30 Mo del: Designer CA 31 Knowledge Mo del Num be r : Designer CA 31 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: Solid Vinyls Context: Select Pro duct Domain T er m Deﬁnition: This is a type of wallpap er that is s old at t he design center. Its a thic k layer of vinyl. Sample Usage: Ja ne felt that so lid vinyl w a llpap er would b e easier to ma in- tain in her k itc hen. 97 4.2.31 Mo del: Designer CA 32 Knowledge Mo del Num be r : Designer CA 32 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #1 Mo del Persp ective: Designer Date: 04/13/ 00 Domain T er m Name: Vinyl Coated Context: Select Pro duct Domain T er m Deﬁnition: This is a type of wallpap er that is s old at t he design center. It has a layer of vinyl ov er the wallpaper . Sample Usage: Jane felt that vin yl coated w allpap er w ould be easier to maintain in her family ro om. 98 4.2.32 Mo del: Designer CA 07 Knowledge Mo del Num be r : Designer CA 07 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n Rep or t #3 Mo del Persp ective: Designer Date: 04/04/ 00 Domain T er m Name: lead-time Related Knowledge Mo dels: Desig ner T A 18 Domain T erm Deﬁnition: Ev er y manufacturer has diﬀer e nt lead-times for diﬀerent pro duct. The lead-time is nothing but the normal delivery time for a pro duct. Sample Usage: The lea d time for the 2- inch woo den blinds that Cynt hia ordered was 4 weeks but still she decided to buy the pr o duct. 99 4.2.33 Mo del: Designer CA 11 Knowledge Mo del Num be r : Designer CA 11 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Price B o o k Context: Generate Quote o r inv o ice Domain T e rm Deﬁnition: Pr ice bo ok consists of price list of all the blinds the design cent er car r ies. It ha s the informa tion of a ll vendors, sp e cial pro ducts they ma ke and measurements fo r the diﬀerent window s izes Sample Usage: Designer r e fer s to the price bo ok to gener ate quo te and to generate inv o ice 100 4.2.34 Mo del: Designer CA 12 Knowledge Mo del Num be r : Designer CA 12 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Headrail Context: Blinds Domain T erm Deﬁnition: Headrail is used to support blinds (Horizon ta l o r vertical) Sample Usage: when customer decides to buy the blinds, the des igner ask the customer if they would like the hea drail to be mounted ins ide the window or outside 101 4.2.35 Mo del: Designer CA 13 Knowledge Mo del Num be r : Designer CA 13 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Slats Context: Blinds Domain T erm Deﬁnition: the individual blinds is called slats when installed vertically Sample Usa g e: Designer pr ovides informa tion to the customer on the size of the s la t, precisely the 1” or 2” depending on the man ufacturer 102 4.2.36 Mo del: Designer CA 14 Knowledge Mo del Num be r : Designer CA 14 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: v anes Context: Blinds Domain T erm Deﬁnition: The individual blind when the blinds are installed horizontally Sample Usag e: Designer ask s the size of the v ane the customer is interested in, it can 1” o r 2” and it dep ends on the manufacturer 103 4.2.37 Mo del: Designer CA 15 Knowledge Mo del Num be r : Designer CA 15 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: V alence Context: Blinds Domain T er m Deﬁnition: V alence is use d to cov er the headrail Sample Usage: Designe r ask s the customer if they would like to buy the v a le nce along with the headr ail a nd also tells them that some manufacturers provide the v alence along with the headrail. 104 4.2.38 Mo del: Designer CA 16 Knowledge Mo del Num be r : Designer CA 16 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: Cords Context: Blinds Domain T erm Deﬁnition: Co n trols pr ovides with blinds to lower or rais e the blinds o r to tilt the blinds Sample Usa ge: Designer asks the customers what kind of controls they ar e int e r ested in, if they w ant the cords tow ards the rig h t o r left for tilting and to low er or raise the blinds. 105 4.2.39 Mo del: Designer CA 17 Knowledge Mo del Num be r : Designer CA 17 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #6 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: W ands Context: Blinds Domain T er m Deﬁnition: Controls provides with blinds to tilt the blinds Sample Usage: Designer asks the custo mers on wha t side they would like the wand to be. 106 4.2.40 Mo del: Designer CA 20 Knowledge Mo del Num be r : Designer CA 20 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #8 Mo del Persp ective: Designer Date: 04/03/ 00 Domain T er m Name: SKU –Sto ck keeping unit Context: Inv oice Domain T e rm Deﬁnition: A num b er to identif y a ll the pr o ducts in c ase of Insto ck pro duct, and a num b er to identify a type of pro duct manufactured by a par ticular vendor Sample Usage: Designer needs to en ter the SKU n umber for each pro duct in the inv o ic e 107 4.2.41 Mo del: Designer CA 21 Knowledge Mo del Num be r : Designer CA 21 Mo del Des c r iption: Domain Dictionar y KA Source: Intro duction, KA Sessio n #8 Mo del Persp ective: Designer Date: 04/03/ 00 108 4.2.42 Mo del: Designer CA 23 Knowledge Mo del Num be r : Designer CA 23 Mo del Des c r iption: Domain Dictionar y KA Source: K A Session #9 and #10 Section 2 .5.9 a nd 2.5.10 Mo del Persp ective: Designer Date: 04/09/ 00 Domain Dictionary Domain T er m Name: Pa y ment Des k Related Knowledge Mo dels: Custo mer T A 37 Domain T er m Deﬁnition: The fr ont desk is a man dr iven counter that accepts payment from the cus - tomer. It has system connected to the inven tory that updates the inv entory after the items have been chec ked out. It also has systems that trig ger the dispatch or ﬁnalize the customer’s request for measurement/installation a nd delivery after the payment has b een made. Sample Usag e: The custo mer pays for the measurements a t the paymen t desk. Domain T er m Name: Recording Sy stem Related Knowledge Mo dels: Desig ner T A 34 Domain T er m Deﬁnition: Recording system is an e le c tronic o r ma nual for m of ﬁling information. The system s aves a nd stor es diﬀerent types of information and o ften ha s the feature of sear ching through the stored infor mation. Sample Usa ge: The desig ner records the measurements on the re cording system. Domain T er m Name: Sample Blinds Related Knowledge Mo dels: Domain T er m Deﬁnition: Blinds which are pr e -measured and are used by the designer to b e g iven aw ay to help the customer in measuring the windo ws and do ors. These blinds give an idea to the customer how to take measur emen ts. Sample Usage: The designer giv es the sa mple blind to th e customer for measurements. 109 Sample Usage: The desig ner gives a sample blind to the cus tomer to use it to take measure men ts. Domain T er m Name: Pa y ment Ce n ter Related Knowledge Mo dels: Custo mer T A 04 Domain T er m Deﬁnition: Place wher e payment is acce pted from the custo mer and a r eceipt genera ted by the system. The payment system usually has infrastruc tur e for accepting and authenticating the credit card and accepting chec ks. 110 Chapter 5 Uniﬁed Kno w ledge Mo del s 5.1 Mo d el: UKM CA 01 Figure 5 .1: Uniﬁed K nowledge Mo del UKM CA 01 111 5.2 Mo d el: UKM CA 02 Figure 5 .2: Uniﬁed K nowledge Mo del UKM CA 02 112 5.3 Mo d el: UKM CA 03 5.3.1 Instructions to measure windo ws Key Co ncepts: 1. T o mea sure the window (width*Leng th) to the nearest 1/ 8th of an inc h. 2. T o mea sure top, middle a nd b ottom of the window 5.3.2 Instructions for horizon tal or cellular blinds Key Co ncepts: 1. T o mea sure the smallest width and the Longest le ngth 2. T o mea sure windows with o r without headra il 5.3.3 Instructions for V ertical Blinds Key Co ncepts: 1. T o mea sure the top w idth and the sma llest length of their window 2. T o mea sure windows with o r without the headrail User: Customer uses the ab ov e information provided by the designer to mea s ure the windows 5.3.4 P ossible controls on Blinds Key Co ncepts: 1. Standard co nt r ols 2. Reverse Standard controls 5.3.5 2” Horizon tal blinds Key Co ncepts: Tilt cord and Lift cor d on right side or left side of the window 113 5.3.6 1” Horizon tal blinds Key concepts: Tilt wand a nd Lift wand on right side or left s ide of the window 5.3.7 Cellular blinds Key Co ncepts: Lift cord on r ight or left side of the window 5.3.8 V ertical Blinds Key concepts: W ands on s ame side as stack or on either side of the window User: Customer uses this info r mation provided by the de s igner to decide on what controls they need for the windows 114 5.4 Mo d el: UKM CA 04 Figure 5 .3: Uniﬁed K nowledge Mo del UKM CA 04 115 5.5 Mo d el: UKM CA 05 Domain Dictionary 5.5.1 Recording system Domain T er m Deﬁnition: Recording s ystem is an electro nic o r manual form of ﬁling information. The system saves and stor es diﬀeren t types o f information and often has the feature o f searching thr ough the stored information. In the design center it is use d to store customer infor mation and pro duct details. 5.5.2 P ayme n t Acceptance System Domain T erm Deﬁnition: The payment des k is a ma n driven counter that ac- cepts payment from the customer. It has system connected to the in ven tory that upda tes the inv entory after the items hav e b een chec ked out. It a lso has systems that tr igger the dispatch or ﬁnalize the customer ’s request for mea surement or installation a nd delivery after the paymen t ha s b een made. The paymen t center generates the receipt after payment has bee n accepted. The paymen t sys tem usually ha s infrastr ucture for accepting and authen tica ting the cr e dit card and accepting checks. 5.5.3 W allpaper bo oks Domain T e rm Deﬁnition: The wallpaper b o oks contain sa mple w allpap er and manuf acturer informatio n. The customer looks at the bo o k to decide the t yp e of pro duct he needs. Sample Usage: The wallpaper b o oks are not so rted by color but you will ﬁnd the colo r you w a n t in every b o ok . The ﬁrst thing the custo mer needs to decide on is the style of wallpaper that a r e con temp or a ry , tra ditional, kitc hen and bath. 5.5.4 Mini Blind- Cutter Related Knowledge Mo dels: None Domain T erm Deﬁnition: The mini blind cutter is used at the design center to cut in sto ck blinds acco rding to c us tomer r e q uirement Sample Usage: Jane w a nt ed to buy in sto ck blinds but the size she wanted was not in sto ck. So Jimm y suggested that she could pic k a la rger size in the 116 brand s he liked a nd that he could cut it to size she needed using the mini blind cutter. 5.5.5 Price B o ok Related Knowledge Mo dels: None Domain T erm Deﬁnition: The price b o ok is the bo ok with all the pro duct sku’s the price and the vendor contact information. Sample Usa ge: When Jane w anted the Bali 1 –inch blinds in 40 by 6 0 and the size was not in sto ck, Jimmy lo oked at the Price b o ok and found the vendor nu m be r and called him up to a s k for the delivery date for that par ticular type. 5.5.6 Catalog Related Knowledge Mo dels: None Domain T erm Deﬁnition: The c atalogs are the product manuf acturer pub- lished magazines that hav e all the diﬀeren t mo dels man ufacture d by the com- pany for the customer to select. Sample Usage: When Cynthia wanted to decide on the diﬀere nt t yp es of the blinds then J ulie directed her tow ar ds the cata logs to see the diﬀerent t yp es catalogs stacked up on the shelf. 5.5.7 Lead t ime Domain T erm Deﬁnition: Every ma nufacturer has diﬀeren t lead-times for dif- ferent pro duct. The lead-time is nothing but the nor mal delivery time for a pro duct. Sample Usage: The lea d time for the 2- inch woo den blinds that Cynt hia ordered was 4 weeks but still she decided to buy the pr o duct. 5.5.8 Headrail Context: Blinds Domain T erm Deﬁnition: Headrail is used to support blinds (Horizon ta l o r vertical) Sample Usage: when customer decides to buy the blinds, the des igner ask the customer if they would like the hea drail to be mounted ins ide the window or outside 117 5.5.9 Slats Context: Blinds Domain T erm Deﬁnition: the individual blinds is called slats when installed vertically Sample Usa g e: Designer pr ovides informa tion to the customer on the size of the s la t, precisely the 1” or 2” depending on the man ufacturer 5.5.10 V anes Context: Blinds Domain T erm Deﬁnition: The individual blind when the blinds are installed horizontally Sample Usag e: Designer ask s the size of the v ane the customer is interested in, it can 1” o r 2” and it dep ends on the manufacturer 5.5.11 V alence Context: Blinds Domain T er m Deﬁnition: V alence is use d to cov er the headrail Sample Usage: Designe r ask s the customer if they would like to buy the v a le nce along with the headr ail a nd also tells them that some manufacturers provide the v alence along with the headrail. 5.5.12 Cords Context: Blinds Domain T erm Deﬁnition: Co n trols pr ovides with blinds to lower or rais e the blinds o r to tilt the blinds Sample Usa ge: Designer asks the customers what kind of controls they ar e int e r ested in, if they w ant the cords tow ards the rig h t o r left for tilting and to low er or raise the blinds. 5.5.13 W ands Context: Blinds Domain T er m Deﬁnition: Controls provides with blinds to tilt the blinds 118 Sample Usage: Designer asks the custo mers on wha t side they would like the wand to be. 5.5.14 SKU –St o c k Keeping Unit Context: In voice Domain T e rm Deﬁnition: A num b er to identif y a ll the pr o ducts in c ase of Insto ck pro duct, and a num b er to identify a type of pro duct manufactured by a par ticular vendor Sample Usage: Designer needs to en ter the SKU n umber for each pro duct in the inv o ic e 5.5.15 Stage for register Context: In voice g eneration Domain T er m Deﬁnition: After all the pro duct information is entered by the designer in to the system, Stag e for reg ister is the command used in the Inv oice generation s o ft ware to generate the Inv oice. Sample Usage: Desig ner g e ne r ates the inv oice for the customer 5.5.16 Measuremen t s Context: Select P r o duct Domain T erm Deﬁnition: Each blind ha s measurements, which are its length and the breadth they ar e in inc hes. These mea surements are used determine ﬁts customer’s window. So when a customer wan ts to buy the pro duct he needs to get the window s ize. Sample Usa ge: Ja ne measured her window fo und that it was 30 X 60 so with these measur ement s she wen t to the design ce n ter to reinstall her blind. 5.5.17 V erticals Context: Select P r o duct Domain T erm Deﬁnition: This is a type of blind that is sold a t the design center. They hav e the individual pieces of the blinds arranged in a vertical pattern. Sample Usage: Ja ne lo o ked at the large window facing the sea and decided that vertical blinds there w o uld give her easy a ccess to the patio. 119 5.5.18 Horizon t al Context: Select P r o duct Domain T erm Deﬁnition: This is a type of blind that is sold a t the design center. They have the individual pieces o f the blinds arra nged in a horizontal pattern. Sample Usage: Jane lo oked at her kitchen window and felt that horizontal blinds ther e would b e e a sy to maintain. 5.5.19 Cellular Context: Select P r o duct Domain T erm Deﬁnition: This is a type of blind that is sold a t the design center. They have the individual pieces o f the blinds arra nged in a horizontal pattern but they give more pr iv a c y than the horiz ontal. Sample Usa ge: J ane lo oked at her b edro om window and felt that cellular blinds ther e would give her eno ugh priv acy . 5.5.20 Em b ossed Context: Select P r o duct Domain T er m Deﬁnition: This is a type of wallpap er that is s old at t he design center. They hav e the puﬀy impress io ns. Sample Usage: Jane felt that em b ossed wallpaper would g ive her mor e e le - gance to her dining r o om 5.5.21 Solid Vin yls Context: Select P r o duct Domain T er m Deﬁnition: This is a type of wallpap er that is s old at t he design center. Its a thic k layer of vinyl Sample Usage: Ja ne felt that so lid vinyl w a llpap er would b e easier to ma in- tain in her k itc hen. 5.5.22 Vin yl Coated Context: Select P r o duct 120 Domain T er m Deﬁnition: This is a type of wallpap er that is s old at t he design center. It has a layer of vinyl ov er the wallpaper . Sample Usage: Jane felt that vin yl coated w allpap er w ould be easier to maintain in her family ro om. 121 Chapter 6 Reference Arc hitecture 6.1 Arc hitecture Ov erview The Domain Refere nc e Archit ecture has four classes . - Customer, Measurer , Installer and Designer. The architecture ha s role based DRA C Resp onsibility assignment. Primary goal of this architecture is to reduce coupling, roundtrip and dep endencies betw een the DRAC. The Deriv ation Pro cess was a lso aimed at max imizing Reusability . Assigning the services to the DRA C that needs it, reduced coupling. Ser vices, which ar e shared b etw een the DRACs, ar e the ser - vices, which are needed by b oth the DRA C. Assig ning da ta/even t to the DRAC that needs it mo st also reduced the ro undtrip. The architecture is indep endent of any technology classes. This ensur es re us ability of the a rchitecture. 122 6.2 Comp onen t Organization Figure 6 .1: Comp onent O r ganization. 123 6.3 Primitiv e Comp onen ts 6.3.1 DRAC : Designer Declarative Mo del: DRA C Name: Designer Created by: Haritha Nandela Domain Reference: Designer UKM T A 13, Designer UKM T A 04 , De- signer UKM T A 15 , Designer UKM T A 1 4, Designer UKM T A 24, Des ig ner UKM T A 06, Designer UKM T A 05, Designer UKM T A 07, Designer UKM T A 12, Designer UKM T A 08, Designer UKM T A 16, Designer UKM T A 18, De- signer UKM T A 17 , Designer UKM T A 1 9, Designer UKM T A 25, Des ig ner UKM T A 26, Designer UKM T A 27, Designer UKM T A 28, Designer UKM T A 29, Description: This DRAC ba sed on the p erfo rmer DESIGNER. Rationale: DRA C Attributes: Name: Measurements Description: E ach blind has measur ement s, whic h are its le ng th and the breadth they are in inches. These measurements are used determine ﬁts cus- tomer’s window. So when a customer wan ts to buy the pro duct he needs to ge t the window size. Cardinality: 1 Domain Mo del Reference Number:UKM CA 04 , UKM CA 05 Name: Customer Infor mation Description: Name, Deliv e r y site address, Ins tallation site address , Phone nu m be r Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: Pro duct Infor mation Description: Pro duct T yp e, Price , SKU, V endo r Name Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: V endor Information Description: Name, Addres s , P hone num b er, F a x Number 124 Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: Pro duct Av ailability Description: Y e s o r no indicating pr o duct av a ilability Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: Customer Requir ement s Description: Price, Range, colo r, who - the decision maker, when- when cus- tomer requir e s, where - where customer want s the pro duct to be installed, wh y - what the customer is trying to achiev e Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: Quote Description: Measurements , P ro duct Informa tion Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: Installation Sta tus Up dated Description: None Cardinality: Even t Domain Mo del Reference Number: UKM T A 01, UKM CA 04 Name: Quote Gener ated Description: None Cardinality: Even t Domain Mo del Reference Number: UKM T A 01, UKM CA 04 Name: Customer infor mation Stor ed Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 Name: Measurements recorded Description: None Cardinality: Even t 125 Domain Mo del Reference Number:UKM T A 0 1, UKM CA 04 Name: Designer no te price from price list into system Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 Name: Designer ha s key e d in the information Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 Name: Stage for Register Description: After all the pro duct information is e ntered by the designer in to the system, Stage for reg ister is the command used in the Inv o ice gene r ation softw are to genera te the Inv o ice. Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 05 Name: Pro duct infor mation rec o rded Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 Name: Generated inv oice co rrect Description: None Cardinality: Even t Domain Mo del Reference Number: UKM CA 04 , UKM T A 01 Name: Pro duct a nd vendor information sent Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 Name: Pro duct a nd Measurements sent Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 126 Name: Blinds c ut Description: None Cardinality: Even t Domain Mo del Reference Number:UKM T A 01, UKM CA 04 DRA C Service s: Service Na me: Co o rdinate with vendor on pro duct av a ilability Domain Reference: Desig ner UKM T A 13 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: V endor Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: V endor Information Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Input Even ts— Input Even t Name: Customer requests quote Received fr o m Ser vice: None Received fr o m Attribute: None Received fr o m DRAC: Customer Output Data — Output Data Name: Pr o duct Av ailability Sent to Service: Generate q uote for 3 bra nds Genera te quo te on av e r age window siz e , Generate quote o n three diﬀerent pro duct types Sent to DRAC: Designer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task 127 Criticality: High Name: V endor Information Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Post Condition— Name :Pro duct Av ailability Post C o ndition: Pro duct Av ailability is requir ed b efore Genera te quote for 3 brands Gener ate quote on av erag e window s ize, Gene r ate quote on three dif- ferent pro duct types Service Na me: Ga ther customer requir e men ts Domain Reference: Desig ner UKM T A 04 Service Dur ation: 1- 2 hours Execution F req uency: Discrete Input Even ts— Input Even t Name: Customer approached Received fr o m Ser vice: None Received fr o m Attribute: None Received fr o m DRAC: Designer Output Data — Output Data Name: Customer requir emen ts Sent to Ser vice: Pr esent Types Sent to DRAC: Designer Service Na me: Gener ate quo te on three diﬀerent br ands Domain Reference: Desig ner UKM T A 15 Service Dur ation: 5- 3 0 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer 128 Input Data Name: Measur ement s Received fr o m Ser vice: Reco rd Meas ur ement s Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Av a ilability Received fr o m Ser vice: Co o rdinate with vendor on pro duct av aila bility Received fr o m Attribute: Pr o duct Av ailability Received fr o m DRAC: Designer Output Data — Output Data Name: Quo te Sent to DRAC: Customer Output E ven t— Output E ven t Name: Quote generated Sent to DRAC: Customer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Present Brands ser vice sho uld b e completed b efor e this task Criticality: High Name: Measurements Pre-co ndition: Present Brands ser vice sho uld b e completed b efor e this task Criticality: High Name :Pro duct Av ailability Post Conditio n: Coordina te with v endo r on pro duct a v a ilability should be completed b efore this task Service Na me: Gener ate quo te on av era ge window size Domain Reference: Desig ner UKM T A 14 Service Dur ation: 5- 1 0 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands 129 Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Input Data Name: Pr o duct Av a ilability Received fr o m Ser vice: Co o rdinate with vendor on pro duct av aila bility Received fr o m Attribute: Pr o duct Av ailability Received fr o m DRAC: Designer Input Even t— Input Even t Name: Customer requests a quote Received fr o m DRAC: Customer Output Data — Output Data Name: Quo te Sent to DRAC: Customer Output E ven t— Output E ven t Name: Quote generated Sent to DRAC: Customer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Present Brands ser vice sho uld b e completed b efor e this task Criticality: High Name :Pro duct Av ailability Pre Condition: Coo rdinate with vendor on pro duct av ailability should b e completed b efore this task Service Na me: Gener ate quo te on three diﬀerent pr o duct types Domain Reference: Desig ner UKM T A 24 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer 130 Input Data Name: Measur ement s Received fr o m Ser vice: Reco rd Meas ur ement s Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Av a ilability Received fr o m Ser vice: Co o rdinate with vendor on pro duct av aila bility Received fr o m Attribute: Pr o duct Av ailability Received fr o m DRAC: Designer Output Data — Output Data Name: Quo te Sent to DRAC: Customer Output E ven t— Output E ven t Name: Quote generated Sent to DRAC: Customer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Present Brands ser vice sho uld b e completed b efor e this task Criticality: High Name: Measurements Pre-co ndition: Present Brands ser vice sho uld b e completed b efor e this task Criticality: High Name :Pro duct Av ailability Post Conditio n: Coordina te with v endo r on pro duct a v a ilability should be completed b efore this task Service Na me: P resent Bra nds Domain Reference: Desig ner UKM T A 06 Service Dur ation: 5- 1 0 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Pr o duct Type Received fr o m Ser vice: P r esent Types 131 Received fr o m Attribute: Pr o duct Type Received fr o m DRAC: Designer Output Data — Output Data Name: Pr o duct Information Sent to Serv ice: Genera te quo te for 3 bra nds, Co ordinate with vendor on pro duct Av aila bilit y , Generate quote on three diﬀerent pro duct types, Notes price list of price list pro duct, enter pro duct infor mation. Se nd Pro duction information to vendor, Send pro duct Infor mation to Installer Sent to DRAC: Designer , Installer Output Data Name: Pr o duct Information Sent to Service: Co ordina te with vendor on pro duct Av aila bilit y , Send Pr o - duction information and vendor to vendor Sent to DRAC: Designer , Installer Pre Condition— Name: Pro duct Types Pre-co ndition: Pres ent Types sho uld b e completed befor e this task Criticality: High Post Condition— Name :Pro duct Infor ma tion Post Condition: Pro duct Informa tio n is obtained when this task is com- pleted. Service Na me: P resent Types Domain Reference: Desig ner UKM T A 05 Service Dur ation: 5 minutes- 1 hours Execution F req uency: Discrete Input Data— Input Data Name: Customer Requirements Received fr o m Ser vice: Gather Customer requirements Received fr o m Attribute: Customer Requirements Received fr o m DRAC: Designer Output Data — Output Data Name: Pr o duct Type 132 Sent to Service: Generate quote for 3 pro duct types, Send re q uest mea surer, Sent to DRAC: Designer Pre Condition— Name: Customer Requir ement s Pre-co ndition: Gather custo mer requirements sho uld be completed b efore this ta sk Criticality: High Post Condition— Name :Pro duct Type Post Condition: P ro duct Type is obtained when this task is co mpleted. Service Na me: Reco rd Customer informatio n Domain Reference: Desig ner UKM T A 07 Service Dur ation: 5- 1 5 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Customer Information Received fr o m Ser vice: None Received fr o m Attribute: None Received fr o m DRAC: Designer Output E ven ts— Output E ven t Name: Customer Information recorded Sent to Ser vice: Send request to meas urer, send request to installer Sent to DRAC: Measur er, Installer Pre Condition— Name: Customer Infor mation Pre-co ndition: Customer infor mation needed to complete this se r vice. Criticality: High Post Condition— Name :Pro duct Infor ma tion Post Condition: Cus tomer Information is recorded when this task is com- pleted. 133 Service Na me: Reco rd Mea s urements Domain Reference: Desig ner UKM T A 12 Service Dur ation: 5- 1 0 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Measur ement s Received fr o m Ser vice: Send measur e men ts to desig ner Received from Attribute: Measur ement s and custo mer information received by desig ner Received fr o m DRAC: Designer Input Data Name: Customer Information Received fr o m Ser vice: Send measur e men ts to desig ner Received from Attribute: Measur ement s and custo mer information received by desig ner Received fr o m DRAC: Designer Output E ven ts— Output E ven t Name: Measur ement s recorded Sent to Ser vice: Send request to meas urer, send request to installer Sent to DRAC: Measur er, Installer Pre Condition— Name: Customer Infor mation Pre-co ndition: Customer infor mation needed to complete this se r vice. Criticality: High Name: Measurements Pre-co ndition: Measur emen ts needed to complete this service. Criticality: High Post Condition— Name: Measurements Post Condition: Mea surements is recor ded when this task is completed. Service Na me: Send req ue s t to mea surer Domain Reference: Desig ner UKM T A 08 134 Service Dur ation: 1- 1 0 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Customer Information Received fr o m Ser vice: Reco rd customer information Received fr o m Attribute: Customer informatio n stored Received fr o m DRAC: Designer Input Data Name: Pr o duct types Received fr o m Ser vice: P r esent types Received fr o m Attribute: Pr o duct type Received fr o m DRAC: Designer Output E ven ts— Output Data Name: Measurer receives customer informa tion and pro duct t yp e infor mation Sent to Ser vice: Check Av ailability with cus tomer Sent to DRAC: Measur er Pre Condition— Name: Customer Infor mation Pre-co ndition: Customer infor mation sho uld b e completed b efor e this task Criticality: High Name: pro duct Typ e Pre-co ndition: Pres ent t yp es should b e completed b efore this task Criticality: High Service Na me: Desig ner notes price list o f selec ted pro duct Domain Reference: Desig ner UKM T A 16 Service Dur ation: 5- 1 5 Minut es Execution F req uency: Discrete Input Data— Input Data Name: Measur ement s Received fr o m Ser vice: Reco rd Meas ur ement s Received fr o m Attribute: Mea s urements r ecorded 135 Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Output E ven t— Output E ven t Name: Designer notes price from price list into sy stem Sent to Ser vice: Enter price information Sent to DRAC: Designer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Name: Measurements Pre-co ndition: Measur emen ts should b e completed b efore this task Criticality: High Service Na me: E nt e r price information Domain Reference: Desig ner UKM T A 18 Service Dur ation: 10- 25 Minutes Execution F req uency: Discrete Input Even ts— Input Even t Name: Entering price list for the selected pr o duct Received fr o m Ser vice: Desig ne r notes price list of the se le c ted pro duct Received fr o m Attribute: Desig ner notes price fro m pr ice list into s ystem Received fr o m DRAC: Designer Output E ven t— Output E ven t Name: Stage for r egister Sent to Ser vice: Gener a te inv oice and ha nd it to customer Sent to DRAC: Designer Pre Condition— Name: Notes price list for the se lected pro duct 136 Pre-co ndition: D esigner notes price list o f selected product sho uld b e com- pleted b efor e this task Criticality: High Post Condition— Name : Stag e for register Post Condition: Stag e for register is co mpleted a t the end of this s ervice. Service Na me: E nt e r pro duct information Domain Reference: Desig ner UKM T A 1 7 Service Dur ation: 10- 25 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Measur ement s Received fr o m Ser vice: Reco rd measur ement s Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Output E ven t— Output E ven t Name: Pro duct information recorde d Sent to Serv ic e: Sent pro duct and measurements to vendor, send request to installer Sent to DRAC: Designer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Name: Measurements Pre-co ndition: Measur emen ts should b e completed b efore this task Criticality: High 137 Post Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Service Na me: Gener ate inv o ice and hand it to customer Domain Reference: Desig ner UKM T A 19 Service Dur ation: 5 Minutes Execution F req uency: Discrete Input Even ts— Input Even t Name: Stage for r e g ister Received fr o m Ser vice: E nter price information Received fr o m Attribute: Stag e for register Received fr o m DRAC: Designer Output Data — Output Data Name: Inv o ice Sent to Ser vice: Pay cashier Sent to DRAC: Customer Pre Condition— Name: Stage for regis ter Pre-co ndition: Stage for r egister should b e co mpleted b efore this task Criticality: High Post Condition— Name : Inv oice Post Condition: Inv oice is required b efore s e rvice Pay cas hier Service Na me: Send pro duct info a nd mea surements to receiving dept Domain Reference: Desig ner UKM T A 25 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: V endor Information Received fr o m Ser vice: P r esent Brands 138 Received fr o m Attribute: V endor Information Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Output E ven t— Output E ven t Name: Pro duct info and vendor info sent Sent to Ser vice: None Sent to DRAC: Exter nal Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Name: V endor Information Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Service Na me: Send pro duct information a nd mea surements to vendor Domain Reference: Desig ner UKM T A 26 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Measur ement s Received fr o m Ser vice: Reco rd measur ement s Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Output E ven ts— 139 Output E ven t Name: Pro duct info and meas urements send to vendor Sent to Ser vice: None Sent to DRAC: Exter nal Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Name: Measurements Pre-co ndition: reco r d measurements sho uld b e completed befor e this task Criticality: High Service Na me: Send req ue s t to ins ta ller Domain Reference: Desig ner UKM T A 27 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Meas urements Rec e ived from Serv ic e: Reco r d Measur- men ts Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Input Data Name: V endor Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: V endor Information Received fr o m DRAC: Designer Output E ven ts— Output E ven t Name: Pro duct and vendor information r eceived Sent to Ser vice: Check av ailability with custo mer Sent to DRAC: Measur er 140 Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Name: Measurements Pre-co ndition: reco r d measurements sho uld b e completed befor e this task Criticality: High Service Na me: Check Inv entory Domain Reference: Desig ner UKM T A 2 8 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Meas urements Rec e ived from Serv ic e: Reco r d Measur- men ts Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Output Data — Output Data Name: Pr o duct Av ailability Sent to Ser vice: Cut Blinds Sent to DRAC: Designer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Post Condition— Name :Pro duct Av ailability Post Condition: P ro duct Av ailability is re q uired b efore we can cut the blinds 141 Service Na me: Cut Blinds Domain Reference: Desig ner UKM T A 29 Service Dur ation: 10- 15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Measur ement s Received fr o m Ser vice: Reco rd Meas ur ement s Received fr o m Attribute: Mea s urements r ecorded Received fr o m DRAC: Designer Input Data Name: Pr o duct Information Received fr o m Ser vice: P r esent Brands Received fr o m Attribute: Pr o duct Informatio n Received fr o m DRAC: Designer Input Data Name: Pr o duct Av a ilability Received fr o m Ser vice: Check Inv entory Received fr o m Attribute: Pr o duct Av ailability Received fr o m DRAC: Designer Output E ven ts— Output Data Name: Blinds cut Sent to Ser vice: Pay cashier Sent to DRAC: Customer Pre Condition— Name: Pro duct Infor mation Pre-co ndition: Pres ent Brands should b e completed b efore this task Criticality: High Name: Measurements Pre-co ndition: Record Mea surements should b e completed b efore this task Criticality: High Name: Pro duct Av ailability Pre-co ndition: Check Inv entory should b e completed b efore this task Criticality: High 142 Post Condition— Name : Blinds Cut Post Condition: Blinds need to be cut b efore the cas hier can be paid. Int e gration Mo del: Dependencies o n o ther DRAC’s: Services Required: None A ttr ibutes Requir ed: A ttr ibute Name: Inv oic e DRA C that owns the a ttribute: Cus tomer A ttr ibute Name: Customer requests quote DRA C that owns the a ttribute: Cus tomer A ttr ibute Name: Customer approa ched DRA C that owns the a ttribute: Cus tomer A ttr ibute Name: Inv oic e handed over to customer DRA C that owns the a ttribute: Cus tomer A ttr ibute Name: Measur ement s and c us tomer infor mation rec e ived DRA C that owns the a ttribute: Mea surer A ttr ibute Name: Measurer r eceives customer infor ma tion and pro duct t yp e information DRA C that owns the a ttribute: Mea surer A ttr ibute Name: Pr o duct information and v endor informatio n received DRA C that owns the a ttribute: Ins taller 6.3.2 DRAC : Measurer Declarative Mo del: DRA C Name: Measur e r Created by: Ansh uma n Sinha Domain Reference: Measurer UKM T A 09, Mea surer UKM T A 22, Mea- surer UKM T A 21, Measurer UKM T A 23, Description: This DRAC ba sed on the p erfo rmer MEASURER. Rationale: 143 DRA C Attributes: Name: Measurements and customer informatio n received. Description: The even t is trigger ed when the measurements and customer information is reco r ded int o the r ecording system. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Av aila bility conﬁrmed. Description: The even t is trigger ed when the av ailability of the customer is conﬁrmed. The customer will b e av a ila ble for when the measure r will visit the site. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Measurer r eceives customer and pr o duct type informa tion. Description: This even t is marked when the measurer re c eives the customer and pro duct type information. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Measurements received. Description: This e vent is ma r ked when the designer rec e ives the meas ure- men ts. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 DRA C Service s Service Na me: Ar rive at site and take measur ement s . Domain Reference: Desig ner UKM T A 11 Service Dur ation: 15-30 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Customer Information Received fr o m Ser vice: Reco rd customer information. Received fr o m Attribute: Event: Cus to mer informa tion stor e d. Received fr o m DRAC: Designer 144 Input Even t Name: Av a ilability co nﬁrmed Received fr o m Ser vice: Check av aila bilit y with customer . Received fr o m Attribute: Av ailability co nﬁrmed Received fr o m DRAC: Measur er. Input Data Name: Pr o duct Type Received fr o m Ser vice: P r esent Types Received fr o m Attribute: Pr o duct Type. Received fr o m DRAC: Designer . Output Data — Output Data Name: Meas urements. Sent to Ser vice: None Sent to DRAC: Measur er. Pre Condition— Name: Customer Infor mation (Data) Pre-co ndition: Customer infor mation sho uld b e av ailable for this service. Criticality: High Name: Customer Av ailability . Pre-co ndition: Customer has to b e av a ilable for the m easurements to be taken. Criticality: High Name: Pro duct Type. Pre-co ndition: Measurer has to know the pro duct type b efore he can take measurements. Criticality: High Post Condition— Name: Measurements Post Co ndition: Measurements are taken and this is the output of the ser vice. Service Na me: Send measur ement s to designer. Domain Reference: Desig ner UKM T A 10 Service Dur ation: 10-15 Minutes Execution F req uency: Discrete 145 Input Data— Input Data Name: Measur ement s. Received fr o m Ser vice: Arr ive at site and take measure men ts. Received fr o m Attribute: Data: Meas urements. Received fr o m DRAC: Measur er. Output E ven t— Output Data Name: Meas urements and customer infor ma tion received. Sent to Ser vice: Reco r d Measurements. Sent to DRAC: Designer . Pre Condition— Name: Measurements (Data) Pre-co ndition: Measur emen ts should b e av ailable for this ser vice. Criticality: High Post Condition— Name: Measurements and customer informatio n received. Post Condition: Me a surements and customer information is received by the designer at the end of the service. Service Na me: Check av aila bilit y with customer . Domain Reference: Desig ner UKM T A 9 Service Dur ation: 10-15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Customer Information Received fr o m Ser vice: Reco rd customer information. Received fr o m Attribute: Event: Cus to mer informa tion stor e d. Received fr o m DRAC: Designer Output E ven t— Output Data Name: Av ailability conﬁrmed. Sent to Ser vice: Arr ive at site and take measure ments. Sent to DRAC: Measur er. Pre Condition— 146 Name: Customer Infor mation (Data) Pre-co ndition: Customer infor mation sho uld b e av ailable for this service. Criticality: High Post Condition— Name: Av aila bility conﬁrmed Post Condition: The av a ilability o f the customer is conﬁrmed at the end of the s e r vice. Int e gration Mo del: Dependencies o n o ther DRAC’s: Services Required: None A ttr ibutes Requir ed: A ttr ibute Name: Measur ement s DRA C that owns the a ttribute: Desig ner A ttr ibute Name: Customer Information DRA C that owns the a ttribute: Desig ner A ttr ibute Name: Pr o duct Information DRA C that owns the a ttribute: Desig ner 6.3.3 DRAC : Installer Declarative Mo del: DRA C Name: Installer Created by: Ansh uma n Sinha Domain Refere nce: Installer UKM T A 09, Installer UKM T A 10, Installer UKM T A 1 1 Description: This DRAC is based on the p erformer INST ALLE R. Rationale: DRA C Attributes: Name: Pro ducts delivered. Description: The event is marked when the pro ducts ar e delivered b y the vendor. This is a precondition to pro duct installatio n. Cardinality: Even t 147 Domain Mo del Reference Number: UKM T A 0 1 Name: Pro ducts Ins ta lled. Description: The ev ent is marked when the pro ducts are installed by the installer to customer’s satisfactio n. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Pro duct a nd vendor information received. Description: This e ven t is ma r ked when the desig ner rece ives the pro duct and vendor infor ma tion ab out the pro duct. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 DRA C Service s Service Na me: Ar rive at site and install pro ducts. Domain Reference: Desig ner UKM T A 22 Service Dur ation: 30-50 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Pr o ducts Deliv ered. Received fr o m Ser vice: Check delivery of pro duct. Received fr o m Attribute: Event: P ro duct delivered. Received fr o m DRAC: Installer Input Even t Name: Av a ilability co nﬁrmed Received fr o m Ser vice: Check av aila bilit y with customer . Received fr o m Attribute: Av ailability co nﬁrmed Received fr o m DRAC: Measur er. Input Data Name: Measur ement s Received fr o m Ser vice: Arr ive at site and take measurments. Received fr o m Attribute: Data: Meas urements. Received fr o m DRAC: Measur er. Output E ven t— Output Data Name: Pr o duct delivered. 148 Sent to Ser vice: Arr ive at site and install pro duct. Sent to DRAC: Installer . Pre Condition— Name: Pro ducts Delivered (Event) Pre-co ndition: Pro ducts s hould b e delivered be fore this s ervice. Criticality: High Name: Av aila bility conﬁrmed. (Event ). Pre-co ndition: Customer av a ilability sho uld b e veriﬁed b efor e the service . Criticality: High Name: Measurements. (Data). Pre-co ndition: Measur emen ts should b e av ailable b e fore the servic e. Criticality: High Post Condition— Name: Pro duct Insta lle d. (Event) Post Condition: P ro duct is installed after this s e rvice. Service Na me: Check delivery of pro duct. Domain Reference: Desig ner UKM T A 21 Service Dur ation: 10-15 Minutes Execution F req uency: Discrete Input Data— Input Data Name: Pr o ducts Information. Received fr o m Ser vice: P r esent brands. Received fr o m Attribute: Data: Pr o duct informatio n. Received fr o m DRAC: Designer Input Data Name: V endor Information. Received fr o m Ser vice: P r esent brands. Received fr o m Attribute: Data: V endor Information. Received fr o m DRAC: Designer . Output E ven t— Output Data Name: Pr o ducts delivered. Sent to Ser vice: Arr ive at site and install pro duct. 149 Sent to DRAC: Installer . Pre Condition— Name: Pro duct Infor mation (Data) Pre-co ndition: Pro ducts Information must b e receiv e d before the ser v ice beg ins.. Criticality: High Name: V endor Information. (Data). Pre-co ndition: Pro ducts Information must b e receiv e d before the ser v ice beg ins.. Criticality: High Post Condition— Name: Pro duct Delivered. (Event ) Post Condition: P ro duct is delivered at the end of the ser v ice. Service Na me: Reco rd Status Domain Reference: Desig ner UKM T A 23 Service Dur ation: 10-15 Minutes Execution F req uency: Discrete Input Even t— Input Data Name: Pr o duct Installed. Received fr o m Ser vice: None. Received fr o m Attribute: None. Received fr o m DRAC: Installer Output E ven t— Output Data Name: Installa tion status upda ted. Sent to Ser vice: The installation status is up dated in the reco rding system. Sent to DRAC: Installer . Pre Condition— Name: Pro duct Insta lle d (Event) Pre-co ndition: Pro ducts is insta lled b efore the service star ts. Criticality: High Post Condition— 150 Name: Installation s tatus up dated. (Even t) Post Condition: The status is recorded in the system. Int e gration Mo del: Dependencies o n o ther DRAC’s: Services Required: Name of even t required: Av ailability c o nﬁrmed Name of Service: Check av ailability with Custo mer Name of the DRA C that owns the s ervice: Measurer A ttr ibutes Requir ed: A ttr ibute Name: Measur ement s DRA C that owns the a ttribute: Desig ner A ttr ibute Name: Customer Information DRA C that owns the a ttribute: Desig ner A ttr ibute Name: Pr o duct Information DRA C that owns the a ttribute: Desig ner A ttr ibute Name: V endor Information DRA C that owns the a ttribute: Desig ner A ttr ibute Name: Av ailability Conﬁr med DRA C that owns the a ttribute: Desig ner 6.3.4 DRAC : Customer Declarative Mo del: DRA C Name: Customer Created by: Ansh uma n Sinha Domain Reference: Measurer UKM T A 09, Mea surer UKM T A 10, Mea- surer UKM T A 11 Description: This DRAC is based on the p erformer CUSTOMER. Rationale: DRA C Attributes: Name: In voice 151 Description: The inv o ice has the details of the o r der and the pro duct infor- mation. It also has the custo mer information a nd the measurements along with the pr ice and the tota l a mount to be rece ived from the c ustomer. Cardinality: 1 Domain Mo del Reference Number: UKM CA 04 Name: Pa yment for Inv oice made. Description: The event is marked when paymen t for Inv oice is made by the customer. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Customer Appro aches cas hier Description: The event is marked when the custo mer g o es upto the payment desk to make a payment . Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Customer Reque s ts Quote Description: If the customer w ants to buy the pr o duct he needs to lo o k at the quo te to decide if the pro duct suit the budget. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: Customer Appro ached Description: The customer is approa ched by the desig ner to ga ther r equire- men ts. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 Name: In voice handed ov er to customer . Description: The In voice is handed b y designer the design center. Cardinality: Even t Domain Mo del Reference Number: UKM T A 01 DRA C Service s Service Na me: Pay Cas hier Domain Reference: Desig ner UKM T A 20 Service Dur ation: 10-15 Minutes 152 Execution F req uency: Discrete Input Data— Input Data Name: Inv o ic e Received fr o m Ser vice: Gener ate Inv o ic e and hand it to customer. Received from Attribute: Data: Inv oice; Even t: In voice handed to customer. Received fr o m DRAC: Designer Input Data Name: Customer Approaches cashier. Received fr o m Ser vice: None Received fr o m Attribute: None Received fr o m DRAC: Customer Input Even ts— Input Even t Name: Customer Approaches c a shier. Received fr o m Ser vice: None Received fr o m Attribute: None Received fr o m DRAC: Customer Output E ven ts— Output Data Name: Paymen t for In voice made. Sent to Ser vice: None Sent to DRAC: Customer Output Data Name: Receipt handed to cus to mer. Sent to Ser vice: None Sent to DRAC: Customer Pre Condition— Name: In voice (Data) Pre-co ndition: Inv oice s hould b e generated b e fo re this servic e. Criticality: High Name: Customer a pproaches cashier . Pre-co ndition: Customer ha s to appro a ch the cashier. Criticality: High Post Condition— Name: Pa yment for Inv oice made. 153 Post Condition: This mar ks the end of the service. Name: Receipt ha nded to customer. Post Condition: The customer is handed the r eceipt after the payment is made to the cashier. 154 Chapter 7 Conclusion Home Depo t is the largest net work of home-pro duct retail shops and is made up of sev era l departments. E ach department needs to k eep track of the pro d- uct av a ilability , pro duct choices a nd o rder ba cklogs a nd g enerate invoice based on the selected product. The pr o ject that we have analyze d, is to help the customer a sso ciates keep track o f the products, know the av a ilability of the customer c hoices , generate in voice, co-ordinate with vendor, measurer and In- staller. Incorp ora ting this Information technology pro duct will tr a nslate into better cus to mer sa tis fa ction and more r even ues for Home Dep ot. In retrosp ect, there were several things we could of have done diﬀerently for this exercise. The interview co uld ha ve b een structur e d a bit more eﬃcient ly . W e notice tha t many of our questions were not useful for o ur ﬁnal mo deling . How ever, we did co ver this deﬁciency by conducting mor e interviews a nd a s many questions. In the end we did get most of the informa tion we need for our Domain Reference Architecture. Our initial mo deling also needed more or g ani- zation b ecause it did to ok us some eﬀort to reﬁne o ur mo del b efore they can b e synthesized into our Domain Reference Architecture. Overall, we did ov erco me most of the s et backs we encountered, thus in our nex t so ftw a re engineering pro ject we ar e conﬁdant that we can do a n even b etter job. Overall, in do ing this e xercise we learned a gre a t deal of wha t automation can and can not accomplish. W e also ga ined exp erie nce in gathering the informa tion we need from experts in diﬀerent discipline. T he exp erience of con verting the gathered knowledge in to a framework for a po ssibly r eal application is very exciting a nd re w arding. 155 Chapter 8 Ac kno wledgemen ts W e would like to thank our professor Dr. Suza nne Barber of Universit y of T e x as a t Austin for the gr eat lea rning at her lecture cla s s on Domain Speciﬁc Soft ware Archit ecture. W e are als o g reatly indebted to a ll employ ees o f the stor e esp ecially Julie who had nu merous le ng th y discuss ions on the speciﬁcs of the working of or de r ing s y stem and its main tenance. She display ed lots o f pa tience in knowledge acquisitions sessio ns. O ur spec ia l a ckno wledgement to teaching assistant for the course who help ed us at v ar io us stages of rep ort submission, correctio n a nd its pickup. 156 Chapter 9 References 1. Domain Sp eciﬁc Soft ware Architecture Lecture Notes by Dr. Suzanne Barb er, Univ e rsity of T exas a t Austin 2. Domain Spec iﬁc Softw are Architecture Pro gram, Soft ware Engineer ing In- stitute at C a rnegie Mellon Universit y - T echnical Rep ort by L TC Erik Mettala and Mar c H. Graham 3. A for mal appro ach to soft ware ar chitecture by Rob ert J. Allen 4. A doma in sp eciﬁc softw a r e a rchitecture for adaptive intelligent sys tems b y Barbar a Hayes-Roth, Karl Pﬂeg er, and Ma rko Ba la banovic 5. A domain sp eciﬁc softw are architecture engineering pro cess outline by Will T r acz, Lo u Coglia nese and Patrick Y oung 6. User-Centered Requirements: The s c enario-ba s ed eng ineering pro cess by Karen McGraw and Ka r an Har bison 7. The scenario based engineering pro cess(SEP ): A user cen ter ed approach for the developmen t o f health care systems by K . Harbinson, L. Burnell, J. Kelly a nd G. Haddo ck 8. Knowledge Acquis itio n for exp ert systems b y A. Hart 9. Thinkerto ys by M. Michalk o 10. Knowledge Acquis itio n in the developmen t of a larg e exp ert s ystem by D. Prera u 157

Domain Specific Software Architecture for Design Center Automation

Original Paper

Comments & Academic Discussion

Leave a Comment

Original Paper

Related Papers

Comments & Academic Discussion

Leave a Comment