The Tamil Nadu Forest Department on Friday launched an artificial intelligence (AI) based early warning system that is designed to save wild elephants from getting hit by moving trains.
The system operates with the support of camera-mounted towers installed on the sides of railway tracks, A and B lines, that pass through reserve forest areas of the Madukkarai forest range of the Coimbatore Forest Division and link Tamil Nadu and Kerala.
The AI-based and machine learning enabled system will detect movement of wild elephants that come near the twin single line or cross the tracks and generate alerts.
Forest Minister M. Mathiventhan, who launched the early warning system, said “it is first of its kind in the country”, covering a “highly vulnerable area” of about 7 km.
“A total of 9,028 instances of elephants entering human habitations were reported in Coimbatore division between 2021 and 2023. In the Madukkarai range, the elephants cross railway tracks as part of their natural movement. Eleven elephants have died in collisions with trains since 2008. The aim of the AI-based system is to prevent such incidents,” he said.
Implemented at a cost of ₹7.24 crore, advanced control and command centre inside the forest, is located around 1 km off Walayar inter-State border.
Supriya Sahu, Additional Chief Secretary, Environment, Climate Change and Forest, said developing the AI-based system was a challenge because elephants are highly intelligent and they have learnt to adapt with traditional control measures such as trench and solar fencing.
She said around 130 trains pass via the A and B railway lines per day and nearly 1,000 elephant crossings are reported on these tracks every year. The Forest Department worked closely with the Railways to implement the project, she said.
The AI-based system has seven towers on B line and five on A line, placed at a distance of 500 metres. Each tower has twin 360 degree rotatable thermal night vision cameras, providing clear visuals for a distance of about 900 metres. Staff posted at the control centre can view these visuals and zoom in or out to verify movements picked up by the thermal imaging.
The system will generate alerts by detecting the animal’s presence in three zones based on the distance from the track, namely yellow (100-150 metres), orange (50-100 metres) and red (0-50 metres). Alerts will be sent to railway officials and forest staff to avert collisions.
Ms. Sahu said that the machine learning feature will enable the system to learn more about animal movements and improve its efficiency.
She added that the department will also include tethered drones, which are powered by battery or power connection from the ground and function round the clock, to cover the sandwich forest area between A and B lines.
“This is to supplement the AI-based system and to cover the dark area between A and B lines. If possible, we will also study if the tethered drones themselves can generate alerts. We are expecting to operationalise these drones within three to four months, with the support of the Tamil Nadu Unmanned Aerial Vehicle Corporation,” she added.
Binomial Solutions Private Limited implemented the project for the Department. Principal Chief Conservator of Forests (PCCF) and Head of Forest Force Subrat Mahopatra, PCCF and Chief Wildlife Warden Srinivas R. Reddy, Additional Principal Chief Conservator of Forests and Director of Green Tamil Nadu Mission Deepak Srivastava, Conservator of Forests and Field Director of Anamalai Tiger Reserve S. Ramasubramanian, District Forest Officer N. Jayaraj and Forest staff were present.
