Guardian Path: Time-Aware Social Visibility Framework for Safe Pedestrian Route Navigation Using Machine Learning

Author(s)	Ms. Afza Ruheen, P Bavithra Matharasi
Country	India
Abstract	Mainstream pedestrian navigation apps optimize for distance or speed and ignore whether a route actually feels safe to walk, even though the same street can be lively at noon and deserted by midnight. This paper presents GuardianPath, a framework that operationalizes Jane Jacobs' concept of natural surveillance using only free OpenStreetMap data. Five time-aware features — guardian proximity, active-hour flag, active POI density, anchor presence, and night penalty — are computed for road nodes in Bengaluru and combined into a rule-based visibility formula that generates 12,000 labelled samples. An XGBoost regressor trained on these samples achieves an R² of 0.9937, and its predictions drive a modified Dijkstra search that produces a safety-optimized route alongside the shortest path, with SHAP explanations for every recommendation. On real routes, the safe path adds only about 2% distance at night while active POI density along the corridor rises nearly fivefold by mid-morning, confirming the system tracks when the city is actually awake.
Keywords	Explainable Artificial Intelligence, Natural Surveillance, OpenStreetMap, Pedestrian Safety, Points of Interest, Safe Route Navigation, SHAP, Smart Cities, Social Visibility, Time-Aware Routing, XGBoost
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 8, Issue 5, September-October 2026
Published On	2026-09-07

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research