Planning Your Journey in Audio: Design and Evaluation of Auditory Route Overviews
Nida Aziz, Tony Stockman, Rebecca Stewart · 2022 · ACM Transactions on Accessible Computing · doi:10.1145/3531529
Summary
This research develops and evaluates an auditory route overview system that enables blind users to preview walking routes before embarking on journeys. The system addresses a critical gap: while many navigation aids support live guidance, few tools allow blind travelers to gain advance knowledge of routes—information sighted people routinely obtain from visual maps. The approximately 44 million blind people worldwide lack easy access to preview tools that could increase confidence, enable route comparison, and support strategic travel decisions. The research followed a user-led design approach across three studies. A Preliminary Survey (15 blind participants) established that auditory route overviews were highly valued, rated 4.8/5 for usefulness—significantly higher than overviews for emails, documents, or websites. A Design Study engaged both expert audio designers (8 participants) and blind users (8 participants) to develop design guidelines through workshops and feedback sessions. Finally, a Usability Study (22 sighted, 6 blind participants) evaluated a prototype using route reconstruction tasks with Lego blocks. The resulting system automatically generates sequential audio from Open Street Maps and GraphHopper APIs, integrating four components: synthetic speech for functional information (distances, directions, street names), auditory icons (AIs) representing points of interest like parks, churches, and shops, earcons forming "auditory brackets" to encapsulate repeating PoIs without overwhelming repetition, and a drone sound (footsteps) indicating progress along the route.
Key findings
The Usability Study demonstrated strong performance in route reconstruction and recognition. Sighted participants placed an average of 10.14 PoIs (~91% accuracy) while blind participants achieved 10.22 (~93%). For recognition tasks, sighted participants correctly identified 91% of PoIs and blind participants 95%. Both groups reconstructed route form accurately, with main errors involving street block lengths rather than overall structure. Critically, performance improved significantly over trials. Reconstruction time decreased by approximately 36% for sighted (12.6 to 8 minutes) and 34% for blind participants (19.2 to 12.8 minutes). Route replays also decreased substantially, indicating learning. One-way ANOVA confirmed these improvements were statistically significant (p = 0.0001 for reconstruction time). The integrated audio design proved effective: blind participants preferred speech for functional information but valued AIs for providing environmental context and forming mental models. Most AIs were immediately recognizable (parks: 100%, libraries: 100%, cinemas: 100%), though some caused confusion—fire station and police station sounds were frequently misrecognized due to similar siren sounds. Auditory brackets representing repeating PoIs were conceptually appreciated but harder to use in practice, with 36% recognition error for sighted and 32% for blind participants. Participants struggled to distinguish opening and closing bracket earcons from the AIs, suggesting this design needs refinement. Customization was unanimously requested—participants wanted control over speech speed, content selection, and volume of individual audio stems.
Relevance
This research fills an important gap in accessible navigation by addressing pre-journey planning rather than just live guidance. The finding that blind users strongly prefer route overviews over other auditory overview applications validates the importance of this specific use case. The work demonstrates that auditory displays can effectively convey complex spatial information when thoughtfully designed. For practitioners, the design guidelines are directly applicable: use speech for critical functional information (directions, distances), auditory icons for environmental/landmark context, limit information density between turns to avoid cognitive overload, and provide extensive customization options. The distinction between what designers prioritized (aesthetic abstract representations) versus what users needed (clear functional information) highlights the essential role of user participation in accessible technology design. The software architecture—combining Open Street Maps, GraphHopper APIs, and automated audio generation—provides a reproducible technical foundation. However, the study acknowledges limitations: the current system is passive (no interactive exploration), auditory brackets require redesign, and some AI sounds need better differentiation. Future work should investigate effects on actual travel decisions and add spatialisation for richer mental model formation.
Tags: blind navigation · auditory display · route planning · auditory icons · earcons · sonification · wayfinding · user-led design · mental maps