India's Ministry of Environment, Forest and Climate Change last week held a national workshop in Dehradun focused on a pressing problem: trains and elephants don't mix. Over two days, government officials, railway authorities and conservation scientists examined what works to stop preventable deaths in the forest.
The numbers tell the story. Around 20 elephants are killed in train accidents every year on average, and 94 elephants have died from train collisions between 2019 and 2025. A December tragedy in Assam highlighted the continuing risk: seven wild Asiatic elephants were killed and a calf injured when a high-speed passenger train collided with a herd crossing tracks in Assam. The train involved, the Rajdhani Express carrying 650 passengers, derailed following the impact but caused no human deaths.
One solution already proving its worth is technology. A year after its implementation, an AI and machine learning-enabled surveillance system designed to prevent elephant deaths on railway tracks has proven successful along the busy interstate route between Palakkad in Kerala and Coimbatore in Tamil Nadu. The system operates in the Walayar-Madukkarai forest range where since 2008, 11 elephants died due to collisions with trains.
A year after commissioning, the results show zero elephant accidents, 5,011 AI-generated alerts, and 2,500 safe elephant crossings. How does it work? AI-powered thermal cameras have been installed on 12 towers, monitored by a command centre staffed by local tribal youth, allowing for round-the-clock surveillance of railway tracks that provide real-time alerts to locomotive drivers and patrol teams.
The Tamil Nadu system remains the flagship proof of concept, but India is moving beyond a single location. Pilot installations have been successfully commissioned in four sections under the North East Frontier Railway, covering 64.03 km of elephant corridors and 141 km of railway block sections in Assam. The system is now being replicated in sensitive railway sections in North Bengal and in parts of Odisha under the East Coast Railway.
The acoustic-based approach works differently. An AI-based Intrusion Detection System, termed 'Gajraj Suraksha', uses optical fibre cables laid along tracks to detect vibrations caused by elephant movement, providing real-time alerts to loco pilots, station masters and control rooms, allowing trains to slow down or halt in time.
Yet the ministry's workshop made clear that technology alone won't solve the problem. The Ministry of Environment, Forest and Climate Change has identified 77 railway stretches across 14 states as priority areas for mitigation measures, selected following joint assessments by Project Elephant, the Wildlife Institute of India, Indian Railways and state forest departments.
The recommended mitigation package includes 503 ramps and level crossings, 72 bridge extensions or modifications, 39 fencing or trenching structures, four exit ramps, 65 underpasses and 22 overpasses. Some stretches have already begun construction, with new railway lines now incorporating wildlife-friendly infrastructure from the design stage.
The challenge facing policymakers is a familiar one: balancing competing needs. India is home to over 60% of the global Asian elephant population, with habitats spread across eastern, northeastern, southern and central regions. Habitat fragmentation, land-use changes, high train speeds, night operations and seasonal elephant movements all contribute to collision risks. Railway expansion, essential for economic development and connectivity, cuts through those same habitats.
The good news is that evidence now supports investment in solutions. The Tamil Nadu system cost money, but it prevented deaths and works on a human scale; the tribal youth monitoring the cameras have jobs, and train drivers across the country could be equipped with similar alert systems. Neither technology nor infrastructure fixes every problem. But despite the forest department's best efforts, elephant incidents could not be significantly reduced with traditional patrols. The workshop suggests a pragmatic way forward: use the tools that work, deploy them where the evidence shows they're needed most, and coordinate between the railways and forest departments to make sure the system actually runs.