A Proxy Stakeholder Approach to Requirements Engineering for Inclusive Navigation

Wayfinding, or the ability to navigate one’s surroundings, is crucial for independent living and requires a complex combination of cognitive abilities, environmental awareness, and technology to manage this successfully. Individuals with cognitive impairment (IwCI) often face significant challenges in learning and navigating their environment. Despite its importance, mainstream navigation technologies are rarely designed with their diverse needs in mind. This study reframes the search for places as a socially distributed task and emphasizes the role of proxy stakeholders, who act on behalf or in coordination with IwCI during navigation. Using a qualitatively led mixed-methods approach, which includes an international survey and a three-stage interview study, we examine the real-world strategies that proxy stakeholders employ to support daily navigation. The findings are synthesized into a set of empirically grounded design recommendations that emphasize customisability, collaborative use, and support for routine-based navigation. Our findings highlight key challenges and adaptive practices, which are synthesized into design recommendations that prioritize customisability, routine-based navigation, and multi-user coordination. By introducing the proxy stakeholder concept into the software engineering literature, we propose a more inclusive approach to requirements elicitation and offer practical guidance for designing navigation technologies that better reflect the complex realities of cognitive support.

💡 Research Summary

The paper tackles a critical gap in the design of navigation technologies for people with cognitive impairments (IwCI) by introducing the concept of “proxy stakeholders” – individuals such as caregivers, support workers, family members, or peers who act on behalf of or in coordination with the primary user during way‑finding tasks. Recognizing that traditional requirements engineering (RE) methods (self‑report interviews, questionnaires, or lab‑based studies) often fail to capture reliable needs from IwCI due to limited metacognitive insight and communication barriers, the authors propose a mixed‑methods approach that foregrounds proxy perspectives throughout the RE lifecycle.

The empirical work consists of an international online survey (312 proxy respondents, 58 IwCI) followed by a three‑stage interview series (45 proxies, 12 IwCI). The survey uncovers common real‑world strategies: pre‑defining routes, relying on routine‑based navigation, and delivering multimodal cues (audio, vibration, visual icons) to reduce cognitive load. The interview phase deepens this picture by documenting how proxies handle unexpected changes (e.g., construction, weather), negotiate shared decision‑making, and compensate for missing accessibility data in mainstream map services.

From these data the authors derive five key insights: (1) IwCI depend heavily on repeatable routines, so systems must support routine capture, storage, and automatic replay; (2) multimodal feedback is essential for error tolerance and reduced mental effort; (3) navigation is inherently collaborative, requiring multi‑user access controls and shared dashboards; (4) detailed pedestrian‑level environmental information (sidewalk quality, curb cuts, lighting) is a prerequisite for safe way‑finding; and (5) proxies possess domain knowledge that cannot be inferred from user‑only elicitation.

To operationalize these insights, the paper outlines a three‑step RE process: (i) identify and map proxy roles; (ii) co‑design interaction scenarios with proxies and users; (iii) structure captured requirements into goals, constraints, and support actions, assigning priorities. This framework explicitly models the social and contextual complexity of navigation for IwCI.

The authors present twelve concrete design recommendations: customizable UI (color, font, voice profiles), routine‑based route saving and auto‑replay, multi‑user coordination with role‑based permissions, error‑tolerant interaction (automatic recovery, suggested detours), offline maps enriched with pedestrian‑specific data, context‑aware alerts (weather, time of day), simultaneous multimodal cue delivery, proxy‑only dashboards showing support logs, training modules for both users and proxies, strict privacy and data transparency, compliance with accessibility standards (WCAG), and an ongoing feedback loop for continuous improvement.

Limitations include a sample skewed toward Australian participants and the absence of a functional prototype evaluated in real‑world settings. Future work is suggested to explore cross‑cultural proxy dynamics, develop AI‑driven personalized routing, and conduct longitudinal usability studies.

In sum, by formally integrating proxy stakeholders into RE, the study demonstrates a viable pathway to more inclusive navigation systems that respect the lived realities of cognitively impaired users. It offers a methodological template that can be extended to other vulnerable user groups, advancing the broader agenda of socially aware, user‑centered software engineering.