Abstract
In car-oriented societies where private vehicles dominate urban travel, promoting cycling as a viable mode of transport remains a major challenge. This paper introduces Movita Ride, a mobile application that supports everyday cyclists through data collection, route analysis, and visualization of cycling behavior. The study uses anonymized user data to generate heatmaps of cycling routes in Belgrade, covering both official infrastructure and informal routes. Infrastructure quality was evaluated using the Safe Cycling Routes Toolkit (SCRT) and GIS-based spatial analysis. The results reveal critical mismatches between existing cycling infrastructure and actual user behavior. Movita Ride supports both cyclists and urban planners by providing evidence-based insights into route safety, perceived risk, and infrastructure performance, contributing to modal shift and cycling policy development in a car-dominated urban environment.
- 85.5%of cyclists report feeling unsafe on Belgrade routes
- 100+ km of cycling routes mapped with safety scores
- 104 cyclists surveyed across the Belgrade cycling community
- 50+ confirmed hazard locations identified through field work
Belgrade's cycling modal share sits below 2% of daily trips, despite growing policy ambitions tied to the EU Green Deal and the UN SDG 11 framework. The barriers are structural: fragmented infrastructure, poor surface quality, dangerous intersections, and a near-complete absence of reliable safety data for route planning.
Movita Ride addresses this data gap directly. The Android application allows cyclists to report infrastructure hazards in under 10 seconds, with GPS location captured automatically. Each report feeds a Route Safety Score (RSS) model that combines expert-mapped GIS baselines with crowdsourced field data producing a continuously updated safety heatmap across both the formal cycling network (routes M1–M8) and the informal network (C1–C3) that riders actually use.
The research combines three evidence sources: a structured survey of 104 active cyclists in Belgrade, a stakeholder interview with the Embassy of Germany (whose representative cycled over 3,000 km through the city), and field hazard identification in cooperation with the cycling association "Ulice za bicikliste." The convergence of all three inputs points to the same structural gap: the absence of real-time, spatially granular safety data.
Platform and methodology
The RSS formula applies a length-weighted average across mapped segments: RSS = Σ(Sᵢ × Lᵢ) / Σ(Lᵢ), where each segment of approximately 500 m carries a baseline score derived from GIS analysis. User reports penalize the segment score dynamically with Traffic Danger and active Roadwork carrying the highest weight and the heatmap updates accordingly for all users.
Development roadmap
- MVP TRL 6 · Now. Android app live on Google Play. Hazard reporting (5 categories), Route Safety Score, GeoJSON safety heatmap, Firebase backend, ML classification API.
- 12 months TRL 7–8. Adaptive ML scoring model, municipal admin dashboard, GeoJSON API for city planning systems, iOS application, 500+ active users target.
- 4–36 months TRL 9. Multi-city deployment across the Western Balkans, SUMP reporting integration, EuroVelo corridor data (routes 6, 11, 13), regional safety network.
Read the full paper.
The complete research methodology, data, scoring model and platform architecture is available via Google Drive. https://drive.google.com/file/d/1TjjEzT1vCmubb49U9qjpPIbvJQ81qewp/view