ROURKELA: A research team of the National Institute of Technology - Rourkela (NIT-R) has developed a smart road safety system using the Internet of Things (IoT) and edge computing technology that can see around blind zones of roads and alert drivers to prevent potential collision.
The team from the department of Electronics and Communication Engineering integrated the advanced technologies with roadside units and small on-site processors, which has made it possible to detect approaching vehicles at sharp road turns or curves in real-time to quickly send audio and visual alerts to drivers. The system uses surveillance cameras, with a computer vision algorithm, installed at strategic locations to detect moving vehicles from live video streams and estimate their speed and distance as they approach a blind corner.
Co-authored by BTech graduate KL Sanjeev Tudu, professors Santos Kumar Das, Umesh Chandra Pati, and Poonam Singh, along with research graduates Goutam Kumar Sahoo and Rashmiranjan Nayak from the department, it has been published in International Journal of Computational Vision and Robotics.
The unique feature of this system is that instead of sending all the recorded data to a distant cloud server, the processing is carried out on edge devices located close to the cameras. This reduces communication delays and ensures that warnings are generated in real time. Another key feature of the studied algorithm is the graphical user interface for traffic management centres to monitor sharp corners.
NIT-R director Prof K Umamaheshwar Rao said, “In a real-world test, the proposed ‘IoT-edge paradigm-enabled vision-based driving assistance’ system successfully detected vehicles in blind zones and accurately estimated their speed and distance.
The algorithms performed well even on devices with limited computational power, making the technology cost-effective and scalable. The system responded immediately when vehicles entered the danger zone, providing timely warnings through audio-visual alarms and notifications.”
The system promises efficient way to enhance road safety, especially in rural and hilly regions, where road geometry and limited infrastructure cause significant risk factors. In this setup, intelligence is built into the roadside infrastructure and does not require complicated vehicle modifications.