Road safety with AI: India’s rise as an economic power over the last few decades has been a remarkable story, with a booming economy allowing more and more people to afford motor vehicles. This surge in vehicle ownership has created a pressing demand for improved infrastructure, and the India has developed the world’s second-largest road network after the US. However, this progress comes with the significant challenge of ensuring the safety of citizens on increasingly crowded roads.
Road deaths and injuries are a global crisis, and more so for India. Road transport, while essential, is also the most complex and hazardous system we encounter daily. Over the years, the rise in road accidents in India has been alarming. According to a MoRTH report, India witnessed 4,61,312 road accidents, resulting in 1,68,491 fatalities, and injuring 4,43,366 individuals in 2022. This represents a worrying increase of 11.9% in accidents, 9.4% in fatalities, and 15.3% in injuries compared with the previous year.
READ | Road safety: India’s tyre industry faces acid test
In mid-July 2024, at the High-level Political Forum on Sustainable Development (HLPF) at the UN Headquarters, Suman Bery, Vice Chairperson of NITI Aayog, represented India. While delivering India’s national statement on reinforcing the 2023 agenda, he said India will achieve its SDG targets well before 2030.
While it is true that India may achieve some of its SDG targets before 2030, two SDGs related to road safety — SDG 3.6 (on halving the number of deaths and injuries from road accidents) and SDG 11.2 (on providing access to safe, affordable, accessible, and sustainable transport systems and improving road safety for all) — are very difficult to achieve by 2030 unless India adopts and scales up innovative, AI-powered best practices across states and districts.
The ministry of road transport and highways (MoRTH) in India has made significant amendments to the Motor Vehicles Act, 1988, through the Motor Vehicles (Amendment) Act, 2019, enacted on September 1, 2019. These amendments aimed to introduce a range of reforms to improve road safety. It was claimed at the time that the effective implementation of these amendments would reduce the rate of road accidents, injuries, and deaths by 30%, but after almost five years, this lofty claim has proven to be unfulfilled.
Scope of AI-powered solutions
The MVA Act, 2019, has a promising provision under section 136A for promoting the electronic enforcement of traffic rules, which remains unimplemented. The Supreme Court Committee on Road Safety is closely monitoring its implementation. In compliance with one of its directions, the Intelligent Transport System (ITS) India Forum organised a global roundtable conference on July 10, 2024. This conference sought to generate suggestions for a comprehensive national ITS policy and to brainstorm solutions for safer, smarter, and cost-effective mobility in India.
The conference discussed how India can leverage rapidly growing technologies such as artificial intelligence, machine learning, the internet of things, sensors, advanced driver assistance systems (ADAS), 5G, cellular vehicle-to-everything (CV2X), and cloud computing to solve transportation problems and optimise them quickly. The discussion involved over 200 global experts from industry, IITs, policymakers from various ministries, and major stakeholders from ITS, infrastructure industries, and academia.
It was highlighted that irresponsible driving behavior is a major cause of accidents, accounting for over 80% of accidents and fatalities. Implementing foolproof, technology-driven solutions can compensate for human mistakes, provide advanced alerts to drivers, and offer actionable insights to enforcement agencies and road operators to save thousands of lives and effectively reduce road accidents to near zero. The conference emphasised the need to advance the development and implementation of ITS to enhance the safety, efficiency, cost-effectiveness, and sustainability of India’s transportation ecosystem in line with the vision of Vikshit Bharat@2047.
The conference proposed several critical suggestions: promoting regional connectivity in the digitalised transport era, developing inclusive transport facilities, integrating different transport modes, encouraging partnerships and scaling up commitments, exploring creative financing mechanisms involving multiple stakeholders, improving the regulatory framework to support public-private partnerships, innovative monitoring and evaluation processes, capacity building of human resources through sharing knowledge, experiences, and best practices, and creating mass awareness about ITS solutions. Additionally, the conference recommended launching world-class M. Tech and MBA programs in ITS in collaboration with top global universities.
Baby steps in AI-enabled technical solutions
Reviewing available best practices using ITS and AI-supported initiatives reveals promising examples for road safety:
AI-powered traffic signals: Cities like Arlington, Dallas, Fort Worth, Garland, and McKinney in Texas, USA, use AI-powered traffic signals to tackle congestion and enhance safety. These cities have implemented an AI software called NoTraffic, which uses sensors and cameras to adjust traffic signals in real time and reduce intersection collisions.
AI-powered sensors in Canada: A 2019 report by researchers at the University of Waterloo highlighted AI-powered sensors built to trigger alarms if children are left unattended in cars. These palm-sized sensors emit radar signals, which bounce back when they encounter people, objects, or animals inside the vehicle.
Intelligent sensor-based traffic signal management system in Germany: Schools in Germany use this AI to ensure student safety while crossing roads. When a student presses the button on a traffic signal, the intelligent sensor detects the number of people wanting to cross and adjusts the duration of the red light for vehicles, ensuring every student crosses safely before the light turns green.
iRASTE in Nagpur: Nagpur city is implementing an AI-based initiative called Intelligent Solutions for Road Safety through Technology and Engineering (iRASTE) to enhance road safety and reduce accidents. Leveraging AI, the project identifies potential accident scenarios and alerts drivers via ADAS and ISA systems. It also monitors ‘gray spots’—areas that could become blackspots if left unaddressed.
AI camera surveillance in Kerala: Kerala seeks to reduce road accidents by half by 2024 and has installed 726 AI cameras to automatically detect traffic violations and send fines. Once the model captures a breach of road rules, the footage goes directly to the central government’s server.
AI-enabled contactless challans in Bengaluru: Bengaluru’s traffic police use AI-enabled cameras to issue contactless challans via SMS for traffic violations. Delhi’s ITMS employs AI and 3D radar-based systems at key junctions to monitor red-light violations and overspeeding.
Intelligent traffic management system (ITMS) in Delhi: The Delhi Police use the 3D radar-based red-light violation detection camera (RLVD) system at 24 junctions to monitor red-light violations and a gantry-mounted radar-based system to detect overspeeding. The system includes high-resolution cameras (7,000-8,000) with sensor-based real-time traffic volume count technology. Each signal has an IP-based public address system for communicating with drivers.
Telangana’s iRASTE project: Led by INAI and IIIT Hyderabad, this initiative applies AI for road safety in public transport, initially covering a few buses with plans to expand. Similarly, Karnataka’s KSRTC is implementing AI-powered Collision Warning Systems and Driver Drowsiness Systems in over 1,000 buses to enhance passenger safety.
The growing use of state-of-the-art technological and AI-based initiatives for safer mobility reflects India’s commitment to leveraging AI to address the grave challenge of ensuring road safety. The preliminary results of these AI-powered models are promising, though scaling them up across states and cities remains challenging.
Road safety: Challenges and way forward
India is preparing itself to use advanced technological and electronic means to address road safety. However, implementing AI-based road safety measures in India presents several challenges:
High cost of technology deployment and maintenance: Advanced AI systems and high-resolution cameras require significant investment, which may be difficult for some municipalities and states to afford. Additionally, skilled personnel are needed to operate and maintain these systems, necessitating ongoing training and education. This challenge can be addressed by increasing investment in AI infrastructure and technology, fostering public-private partnerships, and incentivising innovation in road safety.
Integration with existing infrastructure: Many parts of India still rely on outdated road infrastructure and traffic management systems. Upgrading these systems to support AI technology involves extensive planning, coordination, and convergence.
Establishing a clear regulatory framework: A robust data ecosystem accommodating concerns related to data privacy and security is essential. AI systems collect and process large amounts of traffic, vehicle, and driver data, including personal information. High-quality, real-time data is vital for AI systems to function effectively. This involves standardising data collection processes and encouraging data sharing among different road safety stakeholders, including government agencies, private companies, research institutions, and CSOs for effective participation and convergence.
(Madhu Sudan Sharma is Senior Programme Officer and Shreni Jani is Research Associate at CUTS International, a leading policy research and advocacy group.)