RESOLVE, in close partnership with the Global Tiger Forum, National Tiger Conservation Authority and Clemson University, are excited to announce a major breakthrough in conservation technology: for the first time ever, wild tigers and elephants have been detected by an AI-powered, cryptic camera-alert system, TrailGuard AI, that transmits the images to the cell phones and computers of park managers and other concerned entities in real-time. The elapsed time from the motion sensor triggered by the passing tiger, to loading and performing inference by the AI algorithm on four images, to transmission to the cell network, to the Internet, and to designated authorities is less than 30 seconds, making this technology a true alerting system. To complete the wish list, using cellular communication, the technology can transmit more than 2,500 images on a single battery charge. In short, the dream has come true.

These results are made widely available in a new publication, Mitigating Human-Wildlife Conflict and Monitoring Endangered Tigers Using a Real-time Camera-Based Alert System in the journal BioScience. Since May of 2022 and continuing through the present, TrailGuard AI has been placed in and around five tiger reserves in sections of two of the most productive tiger landscapes in the world: the Kanha-Pench Landscape in Madhya Pradesh, India and the Terai-Arc Landscape in north India and overlapping into lowland Nepal.

This new paper covers the first few months of the field deployment from the monsoon period of 2022. The technology proved its worth almost immediately, detecting tigers moving close to villages and using the same trails as wildlife poachers.

Read the Publication

The new technology was prototyped effectively in Africa for the past four years, before being trialed in India. This tiny AI-embedded camera-alert system runs powerful computer models on-board the camera to weed out false positives before transmission of data, thus saving precious battery life, and sending only what the end user requires. Intel and CVEDIA played a critical role in supplying the necessary computer chips and AI software, respectively, to make this innovation possible. First designed to detect poachers, the camera becomes a wildlife detector simply by updating the AI model to detect a range of endangered wildlife species. A new AI algorithm to be used in India, the rest of Asia, and Africa, detects eight output classes: felids; canids; elephants; rhinos; bears; wild pigs; humans; and a catch-all class of other mammals and large birds. Where cell connectivity is excellent, as in much of India, TrailGuard AI takes advantage of the strong signal. Where there is no cell service, this system transmits images over long-range radio (LoRa) or satellite.

TrailGuard AI Detection Examples

Real-time transmissions of image detection from TrailGuard AI units deployed on a trail in Dudhwa Tiger Reserve, India. A gang of armed poachers used this trail to hunt wildlife in the reserve illegally (left and right), the same trail used by a tiger a few days apart (center). These TrailGuard AI alerts led the authorities to the arrest of these individuals for poaching a few weeks later.

The technology cannot come soon enough. While tiger numbers are in ascendance in India, Nepal, and Bhutan, they are declining in much of the rest of the range. Landscape-scale conservation is the most effective way to save area-sensitive, wide-ranging top predators like tigers. The deployment centered on Kanha-Pench, the most important of the 76 Tiger Conservation Landscapes in this predator's range. The twin anchors of this landscape, Kanha National Park and Pench National Park, and their surrounding habitats, hold more than 500 tigers, the most anywhere. The co-authors of the study are some of the world's leading experts on tigers and tiger conservation.

"TrailGuard AI has moved from proof-of-concept to becoming part of the tool box to reduce human-wildlife conflict. This technology can be spread to the other tiger-range states as it is so easy to deploy."

– Dr. H.S. Negi, former additional chief of wildlife in Madhya Pradesh and Senior Advisor, Global Tiger Forum and senior adviser to Nightjar, the company that produces the system

"We follow the guidelines of the government's directive to 'Make in India.' We are trying to make our technology long-lasting and affordable, and simple to install. You only need to remove the lens cap and, as long as end-users are designated as recipients, they start seeing alerts on their cell phones immediately. This can be a game-changer, to be used by Indian conservationists to protect people and wildlife, and around the world wherever wildlife and human populations interact."

– Piyush Yadav, lead engineer and lead of Nightjar, the company producing TrailGuard AI in India

"As we take TrailGuard AI to scale, this real-time alerting system for parks and wildlife can become a widespread tool to protect endangered nature. Being part of the evolution of this technology has been gratifying."

– Dr. Eric Dinerstein, RESOLVE and Nightjar

Read the full Press Release

Andy Lee