Towards safer journeys: integrating crime awareness and emergency features in a navigation system for Bangladesh

Abstract

Public concern about personal safety during everyday travel in Bangladesh has intensified, particularly in dense urban corridors where incidents such as theft, harassment, and assault shape route choice and mobility confidence. Existing navigation tools optimize for speed and distance but rarely incorporate risk awareness, leaving commuters without actionable, location-specific safety cues. In the prior academic literature, researchers have mapped crime, proposed safety scoring, and explored risk-aware routing; however, most systems remain prototype-level, lack deployment in the Bangladeshi context, and seldom integrate human-centred evaluation or application-grade interfaces. This gap at the application layer motivates new exploration that marries usable design with credible data and transparent methods. In this study, we design and implement a crime-aware mobile navigation application that combines a city-scale crime map, a community “report crime” flow, and a safety-weighted routing module. Our methodology blends HCI and analytics, where a user study (survey and interviews) to surface needs and expectations, a curated Dhaka incident dataset with engineered severity levels, and statistical analyses, for instance, χ2 for association; ANOVA/Kruskal–Wallis for group differences to validate design assumptions about how risk varies across categories. Results indicate a significant association between crime type and gender and coherent separation of engineered severity across crime categories, supporting the product’s hotspot visualization and safety-aware routing choices. We present a working, cross-platform mobile application built with React Native (Expo), FastAPI, and Supabase with crime map, reporting, and safer-route features; an HCI-grounded evaluation of feature usefulness and acceptability; and a statistical analysis that informs design defaults where there are filters, labels, and routing weights. Together, these contributions demonstrate a feasible, user-centred “safety layer” for everyday navigation that can be piloted at a ward scale and extended through community participation.