How far and discernible are public toilets? A city-scale study using spatial analytics and deep learning in Nanjing, China

Everyone needs access to public toilets, yet despite their importance in ensuring timely use, the shortage and limited availability of public toilets remain a global challenge. Conventional assessments of public toilet services often overlook actual usage patterns and focus solely on physical distance, neglecting wayfinding convenience in shaping user experience. This study addresses these gaps by conducting a city-wide analysis of public toilet service efficiency in Nanjing, China, emphasizing both accessibility and wayfinding. An online route planning algorithm was employed to generate a 5-, 10-, and 15-min catchment area map, quantifying the spatial coverage of public toilets. Additionally, deep learning techniques were utilized to detect public toilet signage in Nanjing's old urban area street views to assess wayfinding convenience. The results indicate an average toilet density of 0.91 units per km 2 , with 10-min catchment areas covering 27.47% of the city space, 71.36% of the population, and 82.00% of public service points of interest in Nanjing. However, signage is present at only 38.32% of public toilets within the old urban area, revealing disparities in distribution, coverage gaps, and inadequate signage in certain areas. To address these gaps, Nanjing could expand public toilet facilities to enlarge the 10-min catchment areas and encourage more businesses to open their toilets to the public. Additionally, increasing signage, including for social public toilets, could improve wayfinding and thus enhance the service efficiency of public toilets across the city. This study offers actionable insights for urban planners and policymakers to improve public toilet infrastructure. Moreover, the methodology provides a scalable framework for evaluating urban infrastructure in cities worldwide. • Pioneers city-scale study integrating accessibility and wayfinding for public toilets. • Online route planning and catchment area mapping assess public toilet service efficiency. • Deep learning and online street view maps broaden toilet sign recognition. • Reveals deficiency spatially in Nanjing's public toilets, offering optimization guidance. • Method applicable to evaluating service efficiency of public facilities worldwide.