This work represents a significant step toward improving road safety and preparing for future autonomous vehicle systems in India.
Addressing Vehicle Congestion with AI
The patent, titled “Adaptive Contention Window Optimisation in VANETs using Multi-Agent Deep Reinforcement Learning for Enhanced Performance Model”, was filed by Dr. Arun Kumar, Assistant Professor; Prof. Bibhudatta Sahoo, Professor; and Dr. Lopamudra Hota, Research Graduate, all from the Department of Computer Science & Engineering, NIT Rourkela.
What is VANETs?
VANETs are communication systems where vehicles in close proximity communicate directly with each other. This capability is critical for safety features, such as one car warning others about sudden braking or an obstacle.
The key challenge in VANETs is vehicle overcrowding. When multiple vehicles transmit messages simultaneously, it leads to congestion, which results in delays or lost messages, compromising the system’s functionality.
The NIT Rourkela model offers a solution using multi-agent deep reinforcement learning. This AI system allows each vehicle to adaptively stagger the time of its messages based on the actions of others. Instead of messages competing, the system learns to sequence and prioritize time-sensitive messages. This adaptive adjustment ensures that important alerts are reliably transmitted and reduces the chances of delay.
Driving Safer Mobility in India
Dr. Arun Kumar explained the motivation behind the work, citing the approximately 480,000 road accidents and 172,000 deaths reported in India in 2023, many of which could be prevented using modern technologies. He added that the patent is a small step toward making autonomous vehicles a reality in India, driving the spirit of “Innovate in India” and “Make in India,” he said.
Prof. Bibhudatta Sahoo stated that by addressing potential congestion in VANETs, the findings lay the groundwork for “safer and a more efficient traffic management”.
The enhanced communication model supports various applications, including safety, traffic management and future systems.
