BIGGER PICTURE Nature & food
Elephants on the railway: the seven kilometres watched by AI
A railway through elephant habitat creates a very specific deadline: someone must notice an approaching animal while the train can still respond. Madukkarai puts machine vision into that chain.

- Monitored railway
- 7 km
- Officially reported alerts
- 6,595
- Recorded elephant train deaths
- 0 · Dec 2023–Jan 2026
How the idea works
Watch the crossing zone
Surveillance looks for elephant movement near railway lines.
Pass on the warning
Detected movement alerts railway and forest personnel.
Coordinate the response
People use the warning to manage an encounter before impact.
The railway crosses a living landscape
Two railway lines connecting Tamil Nadu and Kerala pass through the Madukkarai forest range near Coimbatore. India's environment ministry reports eleven elephant deaths there since 2008 before describing the surveillance project. The problem is concentrated enough to target: a vulnerable seven-kilometre stretch, rather than an impossible promise to watch every elephant across India.
Sources: [1]
An alert must reach someone who can act
Tamil Nadu's forest policy describes AI surveillance intended to detect movement even at night and alert locomotive drivers and other responsible staff. That organisational link matters as much as recognition. A detected animal is only an observation; communicating its presence creates an opportunity to change what happens next. The system supports the people coordinating railway and wildlife safety.
Sources: [2]
Thousands of alerts, with a clear end date
In its answer to parliament dated 29 January 2026, the ministry reported 6,595 alerts and 8,589 elephant detections between December 2023 and January 2026, with no recorded elephant deaths from train collisions. Detections are not necessarily distinct animals, and alerts are not equivalent to collisions prevented. Those differences keep a striking operational record intelligible.
Sources: [1]
A useful result without a universal guarantee
This is an official operational account, not a randomized evaluation isolating AI from staff, railway procedures or other protections. State policy places technology among broader coexistence measures, including patrols and habitat management. The inspiring part is the practical combination: concentrate observation where encounters happen, route it to decision-makers, and record outcomes over a stated period.
The original sources
- India Ministry of Environment, Forest and Climate ChangeRajya Sabha Question 71: AI surveillance system to monitor elephant movement
- Government of Tamil NaduForests Policy Note 2025–2026, section 4.10
- Government of Tamil NaduForests Policy Note 2024–2025