Wildlife management is undergoing a transformative shift with the integration of Artificial Intelligence (AI) technology. One of the systems in use is TrailGuard AI, an advanced camera-alert system that identifies wildlife and poachers and sends real-time images to the authorities overseeing key wildlife areas in India. It has captured and relayed images of tigers, in particular, thus proving effective in the detection and prevention of poaching efforts.

A study titled ‘Mitigating human-wildlife conflict and monitoring endangered tigers using a real-time camera-based alert system’ published in the peer-reviewed wildlife journal BioScience in September last year showed how the system ensures swift communication as notifications of tiger sightings are transmitted to a smartphone app within 30 seconds of the camera being activated.

The study was conducted by 15 national and international wildlife biologists and experts, including members of the Global Tiger Forum, Wildlife Institute of India, Dehradun, forest officials from Kanha-Pench wildlife corridor and Dudhwa Tiger Reserve, Clemson University, USA, RESOLVE, a Washington DC-based non-profit organisation, and Nightjar, a social benefit enterprise.

“Compared to traditional camera traps, TrailGuard AI and its features including its design, durability, long battery life, and connectivity from anywhere, offer wildlife conservationists and local communities a more reliable, efficient, and autonomous solution. It can be thought of as a burglar alarm system designed for sanctuaries and reserves or to protect village lands in its buffer,” said Andy Lee, programme coordinator and research scientist, RESOLVE.

Dr Himmat Singh Negi, former inspector general of forests, National Tiger Conservation Authority and current senior advisor to Nightjar, said, “We realised that early alerts from TrailGuard AI by detecting tigers and elephants in real-time in buffer areas, corridors and even multi-use landscapes could promote coexistence. This technology can contribute to the global goal of the rewilding of endangered populations, preventing poaching and illegal tree felling.”

Tools like TrailGuard AI will gain more centrality in wildlife conservation as time goes by, Dr SP Yadav, additional director general (Project Tiger), ministry of environment, forests and climate change and member secretary, National Tiger Conservation Authority (NTCA) said. “Its benefits are threefold: Firstly, it acts as an early warning system, addressing issues within and around the peripheries of protected areas; secondly, it serves as a dedicated anti-poaching mechanism, harnessing technology for enhanced security; and thirdly, it offers a means to monitor natural landscapes where traditional field patrolling is challenging, providing real-time insights.”

The resurgence of the tiger population, a symbol of India’s conservation success, has ironically heightened the risks of conflicts and poaching, especially outside protected reserves. According to the Union environment ministry, the tiger population has increased from 2,226 in 2014 to 3,682 in 2022. Similarly, the elephant population has seen an increase from an estimated range of 27,669-27,719 in 2007 to 29,964 by 2017.

Non-profit group Wildlife Protection Society of India documented 206 tiger deaths, of which 56 cases were related to poaching and seizures in 2023. In fact, tiger mortality peaked last year compared to the previous nine. Not all of the deaths were caused by poaching, to be sure, a majority in most years were on account of natural causes like old age.

Read Part 1 here: How to save the life of a tiger

There is not a single tiger reserve among India’s 54 designated tiger reserves large enough to hold a viable population of tigers, explained Dr Eric Dinerstein, chief executive officer, Nightjar. “The same condition is likely true for Asian elephants. So for both species, wildlife officials must manage tigers and elephants as meta-populations, that is, populations linked by dispersal. India shares wildlife corridors with its neighbouring countries (Nepal, Bhutan, Myanmar, Bangladesh). Often, maintaining these corridors is as important a conservation goal as protecting the parks themselves. In these situations, TrailGuard AI can play another useful role to ensure that the corridors are being used by tigers and elephants, and document the intensity of use and when individuals move between reserves.”

The study team deployed 12 TrailGuard AI camera-alert systems simultaneously from May to July 2022 in the Kanha–Pench corridor in Madhya Pradesh (the corridor connects the Kanha–Pench–Achanakmar population block, which, at over 300 tigers, is the largest tiger population in central India). All TrailGuard AI units transmitted images in the Kanha–Pench corridor study, 8 of 12 detected and transmitted images of tigers, according to the study.

“The TrailGuard AI system detected tigers in 61 trigger events (image captures) with high accuracy. Researchers and forest department officials received notifications of tiger presence via email or push notification between 30 and 42 seconds after detection. These notifications constitute the first-ever transmission of wild tiger detections using embedded AI. These notifications included tiger detections from three units that were within 300 metres of a village and from which there were also daily notifications of villagers grazing cattle or collecting forest products,” the study said.

Seven camera-alert systems in the Dudhwa Tiger Reserve in Uttar Pradesh from early September 2022 to mid-December 2022. “At Dudhwa, one unit within Kishanpur Wildlife Sanctuary captured a group of poachers carrying a gun and knives on two separate dates and a presumably poached chital on the second detection. In addition, one system was positioned near a recent livestock kill to test the use case of monitoring for people potentially poisoning carcasses,” said Negi, adding that authorities were able to quickly identify the individuals via the transmitted images.

The team worked with local forest rangers to identify crossover areas exhibiting recent tiger activity and offering cell coverage. “TrailGuard AI units were deployed in these locations within one kilometre of the forest-agriculture interface primarily on trails and roads that led directly into villages,” said Piyush Yadav, conservation technology fellow at RESOLVE.

“The technology has been successful in detecting poachers in 12 countries, including the likes of the Serengeti, detecting and leading to the arrest of 50 poachers from 20 different poaching gangs. These results better assure the safety of forest rangers,” said Steve Gulick, chief innovation officer, Nightjar and inventor of TrailGuard.

Ramesh Krishnamurthy, who was part of this study and is a scientist specialising in human-wildlife conflict at the Wildlife Institute of India (WII), said there were challenges integrating TrailGuard AI into wildlife management. As with all Artificial Intelligence-based solutions, the assistance is only as good as its training data set. “This requires extensive data collection across various situations and regions, a process that can be resource-intensive and time-consuming,” said Krishnamurthy. He highlights the strategic placement of camera traps as an early warning system, enhancing forest rangers’ effectiveness against poaching.

Dinerstein argued that only the species involved in human-wildlife conflict need to be detected by technology, which is about 16 species globally. Community involvement is also a significant challenge. “The general suspicion towards technology and its potential impact on local communities makes widespread acceptance a slow process. While AI technology like TrailGuard AI can serve as an effective early warning tool, fostering community involvement is crucial for its success.

Krishnamurthy cautions that technology should be viewed as a complementary tool rather than a complete solution. “The synergy between human capacity and technological advancement is key. Striking a balance between technological advancement and the enhancement of human skills is essential for the effective management of wildlife and the success of such technologies,” he added.

The successful implementation of any new technology in wildlife management, particularly systems like TrailGuard AI, hinges significantly on the involvement and adoption by field personnel and local communities. To facilitate this, comprehensive education modules were conducted for the forest department divisions of the Kanha–Pench corridor and local villagers, the study highlighted.

“Community stewards engaged in participatory monitoring in stewardship programs that facilitated community-led responses to events or emergency issues,” said Hrishita Negi, PhD candidate from Clemson University, who was part of the study.

She added that communities that are empowered, trained, and well-equipped can effectively manage human-wildlife conflict at the local level. “In addition, we also provide a multi-element education module to the community members who live within tiger conflict zones.”

ALSO READ: The exciting world of wildlife CSI, and its absolute necessity

The training focused on how to effectively involve village communities in responding to notifications to prevent the poisoning of livestock carcasses and waterholes, quickly alerting villagers about the presence of tigers, dispersing crowds to prevent potential harm to predators, and monitoring livestock grazing in areas frequented by tigers.

Harnessing the power of technology in conservation efforts is more than a necessity, said Virendra Tiwari, director, WII Dehradun. “The aim is to create a sustainable model, ensuring the safety and welfare of both communities and wildlife.”