NIT Rourkela researchers develop smart system to detect vehicles at blind corners

Researchers at the National Institute of Technology Rourkela have developed a smart roadside system capable of detecting vehicles at blind corners in real time and alerting drivers to prevent potential collisions.

The innovation, proposed by a team from the Department of Electronics and Communication Engineering, leverages Internet of Things (IoT) and edge computing technologies to “see” around sharp turns. The framework integrates surveillance cameras, roadside units, and on-site processors to identify approaching vehicles through computer vision algorithms and estimate their speed and distance. Audio and visual alerts are then generated for drivers nearing the blind spot.

Unlike conventional systems that rely on distant cloud servers, the NIT Rourkela model processes data on edge devices located near the cameras. This significantly reduces communication delays and enables real-time warnings. The system also features a graphical user interface for traffic management centres to monitor high-risk intersections.

The research has been published in the International Journal of Computational Vision and Robotics. The study was co-authored by BTech graduate K.L. Sanjeev Tudu along with faculty members Prof. Santos Kumar Das, Prof. Umesh Chandra Pati and Prof. Poonam Singh, and research graduates Dr. Goutam Kumar Sahoo and Dr. Rashmiranjan Nayak.

Highlighting the institute’s growing research output, NIT Rourkela Director Prof. K. Umamaheshwar Rao said it was encouraging to see undergraduate students contributing to impactful international publications.

The team said field tests demonstrated accurate detection of vehicles in blind zones, even on devices with limited computational capacity, making the system cost-effective and scalable. Estimated to cost around Rs 6,000 as a one-time setup, the technology is particularly suited for rural and hilly regions with limited infrastructure.

Also Read: Miracle in ICU! Woman Doctor revives female patient after 30-minute CPR effort, watch

In a related development, the researchers have secured a patent (No. 580396) for a low-cost IMU/GPS-based real-time accident detection and alert system using ESP32. The patented system combines an Inertial Measurement Unit and GPS sensors to automatically detect crashes and alert emergency responders. A companion smartphone application, “Track,” enables victims to quickly connect with emergency contacts and first responders.

The team is now working on a hybrid AI framework that integrates roadside vision systems with embedded accident detection units to build a distributed safety network for smart cities and connected mobility ecosystems.