I

Chennai, May 25 Indian Institute of Technology Madras and Harvard University Researchers have developed a novel Machine Learning (ML) algorithm named ‘CombSGPO’ (Combined Security Game Policy Optimization) that can help in saving wildlife from poaching.

The researchers found that combined and coordinated use of Forest Rangers and drones were the best way to protect wildlife from poaching. As the resources (Rangers and drones) are limited, the researchers developed this algorithm which provides highly efficient strategies that are more scalable than the earlier ones created for the same purpose, says a release from the institute.

The algorithm works by handling resource allocation and strategising patrolling after the extent of resources available had been identified. For this task, it utilises data on the animal population in the conserved area and assumes that poachers are aware about the patrolling being done at various sites.

Balaraman Ravindran, Mindtree Faculty Fellow and Professor, Department of Computer Science and Engineering, IIT Madras, and the Head of Robert Bosch Centre for Data Science and Artificial Intelligence (RBCDSAI), IIT Madras, collaborated with Milind Tambe’s Research Group – Teamcore - at Harvard University, US, to carry out this study. The work has been peer-reviewed and was well received at the 20th International Conference on Autonomous Agents and Multi-Agent Systems.

Ravindran isaid that the work was motivated by the need to perform strategic resource allocation and patrolling in green security domains to prevent illegal activities such as wildlife poaching, illegal logging and illegal fishing. The resources considered are human patrollers (forest rangers) and surveillance drones, which have object detectors mounted on them for animals and poachers and can perform strategic signalling and communicate with each other as well as the human patrollers.

Game theory model

This developed algorithm utilises a Game Theory-based model created by the researchers. (Game theory is a theoretical framework for conceiving social situations among competing players.) In the context of wildlife protection, Game Theory pertains to predicting the areas where poaching may take place. These predictions are based on the earlier poaching incidents and the interaction between poachers and defenders.

Elaborating on this project, Aravind Venugopal, first author of the study, and a Post-Baccalaureate Fellow, RBCDSAI, IIT Madras, said, “The game model and the kind of resources used to simulate such a ‘poaching game’ between the defender (Forest Rangers and drones) and attackers (poachers) are based on the widely-studied ‘Stackelberg Security Game Model’ and are linked to drones that have already been deployed by Air Shepherd (a foundation that deploys drones to stop elephant and Rhino poaching in Africa).”

To extend this research for application in domains such as security, search and rescue and aerial mapping for agriculture among others, the team is trying sample-efficient multi-agent reinforcement learning with the least amount of data since data collection is costly in a real-world scenario, the release added.