Public transport challenges and technology-assisted accessibility for visually impaired elderly residents in urban environments

TL;DR

Problem: navigating urban public transport is difficult for visually impaired elderly in historic cities like Edinburgh. Approach: a mixed-methods study combines TfE live data analysis with semi-structured interviews to assess network structure and user perceptions of AI-assisted navigation; methods include Nearest Neighbour Index ($R=0.172$), Kernel Density Estimation (KDE), and $k$-means clustering ($K=4$). Contributions: quantifies stop clustering, maps usage density, and integrates qualitative themes on independence, memory, and AI readiness. Significance: informs adaptive, multimodal navigation design and policy for inclusive mobility in urban contexts.

Abstract

Independent navigation is a core aspect of maintaining social participation and individual health for vulnerable populations. While historic cities such as Edinburgh, as the capital of Scotland, often feature well-established public transport systems, urban accessibility challenges remain and are exacerbated by a complex landscape, especially for groups with multiple vulnerabilities such as the blind elderly. With limited research examining how real-time data feeds and developments in artificial intelligence can enhance navigation aids, we address this gap through a mixed-methods approach. Our work combines statistical and machine learning techniques, with a focus on spatial analysis to investigate network coverage, service patterns, and density through live Transport for Edinburgh data, with a qualitative thematic analysis of semi-structured interviews with the mentioned target group. The results demonstrate the highly centralised nature of the city's transport system, the significance of memory-based navigation, and the lack of travel information in usable formats. We also find that participants already use navigation technology to varying degrees and express a willingness to adopt artificial intelligence. Our analysis highlights the importance of dynamic tools in terms of sensory and cognitive needs to meaningfully improve independent travel.

Figures (3)

• Figure 1: Frequency distribution of nearest-neighbour distances between all stops (bus and tram) across the Edinburgh transport system. The overlaid smoothed line represents a kernel density estimate (KDE).
• Figure 2: KDE for transport stop density in Edinburgh. The left-hand panel shows the Greater Edinburgh Area, while the right-hand panel depicts a zoomed-in view of the City Centre. Key locations for transport and public life are marked separately.
• Figure 3: Cluster analysis results for geographical locations and usage frequency (activity) of stops. The left-hand panel shows a boxplot representing the distribution of activity within each cluster, measured by the number of vehicles frequenting each stop. An extreme outlier in Cluster 2 with a value of 8,708 vehicles is excluded from the plot, corresponding to the central area where many key interchange stops are located. The upper right-hand panel depicts the clustering of transport stops in the Greater Edinburgh Area based on both location and activity, while the lower right-hand panel shows the same for the City Centre.