Smart Railway Obstruction Detection System using IoT and Computer Vision

2026-05-07

AI Summary

This study addresses the critical safety challenge posed by railway track intrusions—such as wildlife or human-made obstacles—for which existing detection systems suffer from high costs and excessive false alarms, hindering large-scale deployment. To overcome these limitations, this work proposes NETRA, a low-cost, off-grid edge intelligence system that introduces a novel probabilistic fusion mechanism with adjustable thresholds to coordinate passive infrared (PIR) and ultrasonic sensors in triggering the camera, thereby reducing false positives and cutting unnecessary image processing by 52%. By integrating lightweight MobileNet-SSD and YOLOv5 ONNX models, NETRA achieves unified object detection on a Raspberry Pi. Evaluated over 113 intrusion events, the system attains 95% detection accuracy with zero false alarms, an F1-score of 83.5% for elephant identification, 100% alert delivery via LoRa (868 MHz) within 1–2 km, an end-to-end latency of only 2.4 seconds, and a 75% reduction in deployment cost.

Abstract

Railway track intrusions pose a critical safety challenge for Indian Railways, encompassing wildlife incursions and deliberate malicious obstructions. The December 2025 collision in Assam, in which seven elephants were killed by the Rajdhani Express, underscores the urgency of effective real-time detection. Existing solutions such as the optical fiber-based Gajraj system suffer from prohibitive costs (\$1000/km) and high false alarm rates, limiting deployment to only 20 of India's 101 elephant corridors. This paper proposes NETRA, a cost-effective, internet-independent intrusion detection system deployed on Raspberry Pi Zero W and Raspberry Pi 4 edge platforms. NETRA employs probabilistic sensor fusion integrating a PIR motion sensor and an HC-SR04 ultrasonic distance sensor with a tunable threshold (tau_c = 0.65), enabling event-driven camera activation that reduces unnecessary visual processing by 52%. Upon confirmed intrusion, edge-AI classification using MobileNet-SSD (Pi Zero) or YOLOv5 ONNX (Pi 4) identifies threats including humans, large animals, and track obstructions. Confirmed threats are transmitted via LoRa (868 MHz) to alert the locomotive driver within 2.4 seconds end-to-end. Experimental evaluation across 113 motion events demonstrated 95% detection accuracy with zero false alarms through probabilistic fusion, compared to 85% for binary methods. Raspberry Pi 4 with YOLOv5 achieved 83.5% elephant F1-score, a 5.6x improvement over Pi Zero's heuristic approach (14.8%). LoRa communication achieved 100% packet delivery across 1-2 km in field trials. NETRA reduces deployment cost by 75% (\$247/km vs \$1000/km for Gajraj) while providing unified detection of both wildlife and obstruction threats.

Problem

• railway obstruction
• wildlife intrusion
• real-time detection
• track safety
• false alarms

Innovation

• probabilistic sensor fusion
• edge-AI classification
• LoRa communication
• cost-effective railway safety
• wildlife intrusion detection