Brands
Discover
Events
Newsletter
More

Follow Us

twitterfacebookinstagramyoutube
Youtstory

Brands

Resources

Stories

General

In-Depth

Announcement

Reports

News

Funding

Startup Sectors

Women in tech

Sportstech

Agritech

E-Commerce

Education

Lifestyle

Entertainment

Art & Culture

Travel & Leisure

Curtain Raiser

Wine and Food

YSTV

ADVERTISEMENT
Advertise with us

AI system seeks to help prevent train-elephant collisions, but experts raise questions

The Intrusion Detection System deployed by Northeast Frontier Railways shows promise in reducing train-animal collisions but conservationists highlight gaps that need to be addressed for consistency and effectiveness.

AI system seeks to help prevent train-elephant collisions, but experts raise questions

Friday November 08, 2024 , 5 min Read

On October 16, a huge collision was reportedly averted when loco pilots JD Das and his assistant Umesh Kumar of Kamrup Express saw a herd of elephants crossing the railway tracks between Hawaipur and Lamsakhang stations in Assam at 8.30 pm. 

The train, which runs between Howrah and Dibrugarh, was enroute from Guwahati to Lumding. On seeing the elephants, the drivers applied the emergency brakes.

The technology attributed for this narrow escape is the Intrusion Detection System (IDS) called Gajraj Suraksha—adopted by the Northeast Frontier Railway (NFR) in December 2022. It is said to have alerted the pilots of the herd of elephants crossing the track, prompting them to apply the emergency brakes to stop the collision in time.  

This is a notable advancement in mitigating human-animal conflict in a state where, from 2012 to 2022, a total of 30 elephants have died from being hit by speeding trains, according to NFR data.

According to the Ministry of Environment, Forest and Climate Change, a total of 186 elephants were killed after being hit by trains across India between 2009-10 and 2020-21.

Analysing the Railway Ministry’s data, a report by News18 said that, on an average, 15 elephants were killed on railway tracks each year between 2014 and 2022, and in total at least 135 jumbos have died following train hits across India. Of this, the Northeast Frontier recorded the highest number of deaths at 65 during this period.

The Gajraj Suraksha System

According to the Railway Ministry, the IDS technology was developed by the railways in collaboration with some startups, and introduced on a 150-km stretch in Assam two years ago. The implementation cost of the project on the 700-km tracks was ₹181 crore, according to Railways Minister Ashwini Vaishnaw.

 

After the pilot project on IDS, undertaken in the Chalsa- Hasimara section of the Dooars area under Alipurduar division in West Bengal and the Lanka-Hawaipur section under Lumding division in Assam, NFR signed a memorandum of understanding with RailTel Corporation of India Ltd, for installation of IDS in all elephant corridors in the Northeast Frontier region.

The technology is said to also help detect rail track fractures, landslides, and unauthorised activity on high-traffic wildlife corridors​. It was introduced in 11 elephant corridors: five in Alipurduar division and six in Lumding division.

Whenever an elephant steps onto the railway track, the AI-based system issues alerts to train controllers, station masters, drivers, and other relevant parties, prompting them to take immediate safety measures. 

The technology uses optical fibre cables, which are laid beneath the tracks for telecommunications and signalling. Intrusion detection sensors within this optical fibre network sense the vibrations caused by an elephant’s movement on the track, sending real-time alerts to the control room. The system can detect elephants within 5 metres of the cable, thus aiding in collision prevention.

According to Manoj Kumar, the divisional signal and telecommunication engineer at Lumding, on October 16, the optical fibre cables detected the movement of elephants on the tracks based on parameters such as sound, frequency, amplitude, speed and direction using AI (artificial intelligence), and alerted the pilots 80 metres ahead of the herd. This, he says, prompted them to use the brakes in time to avoid collision. 

Conservationists raise questions

While the railways department cites a reduction in elephant fatalities due to this technology, conservationists working with elephants in the region say accidents continue nevertheless. 

An adult bull elephant was killed by Tinsukia-Naharlagun Express at Ranga de-reserve of Dorpong-Doimukh Elephant Corridor in Arunachal Pradesh, as recently as October 22.

The conservationists, who did not wish to be named, have also raised concerns about the technology’s effectiveness and limitations.

A prominent elephant conservationist in Assam observes that, while the IDS has potential, it relies heavily on loco pilots’ quick reactions rather than a fully automated detection and response. 

“It may fail to send accurate alerts, occasionally missing elephants or even setting off false alarms,” he elaborates. 

Another limitation, he adds, is the system’s proximity range of just 50-150 metres from the tracks; hence there may not be sufficient time for the brakes to be applied and the ground forces to clear the animals from the tracks before an oncoming train nears them.

He, along with other conservationists, also go on to say that the accident that was averted on October 16 was quite likely a result of the pilots’ quick reflexes rather than timely alerts by the IDS.

To address these issues, conservationists suggest extending the detection distance to at least 500 meters, which would give forest and railway teams sufficient time to coordinate and take action.

Furthermore, they emphasise that the effective use of this technology depends on consultations with stakeholders, including local conservation experts and forest authorities, to refine and standardise the system. 

Conservationists also stress that collaborative efforts are essential to make wildlife protection along railway corridors in India’s northeastern states truly effective.


Edited by Swetha Kannan