Researchers from Indian Institute of Technology Madras and Harvard University have developed a new machine learning algorithm named “CombSGPO” (Combined Security Game Policy Optimization) that can help save wildlife from poaching.

Researchers have found that the combined and coordinated use of rangers and drones is a good way to protect wildlife from poaching. Resources (Rangers and drones) being limited, the researchers developed this algorithm which provides a good strategy to protect wildlife with the available resources. This new algorithm provides highly effective strategies that are more scalable than previous ones created for the same purpose.

The algorithm works by managing resource allocation and building patrol strategies once the extent of available resources has been identified. For this task, he uses data on the animal population in the conserved area and assumes that poachers are aware of patrols carried out at various sites.

Prof. Balaraman Ravindran, Mindtree Faculty Member and Professor, Department of Computer Science and Engineering, IIT Madras, and Director, Robert Bosch Center for Data Science and Artificial Intelligence (RBCDSAI), IIT Madras, collaborated with Professor Milind Tambe’s research group – Teamcore – at Harvard University, USA to carry out this study.

This developed algorithm uses a model based on game theory created by the researchers. (Game theory is a theoretical framework for designing social situations between competing actors.) In the context of wildlife protection, game theory is concerned with predicting areas where poaching may occur. These predictions are based on past poaching incidents and the interaction between poachers and defenders.

According to the World Wide Fund for Nature (WWF), the wildlife trade is the second greatest direct threat to the survival of species after habitat destruction. While several organizations and regulatory authorities are trying to curb the incidences of poaching, poachers seem to have always been one step ahead of patrollers. This collaborative research work carried out by two reputable universities will help control poaching incidents.

To extend this research for application in areas such as security, search and rescue, and aerial mapping for agriculture, among others, the team is attempting to perform efficient multi-agent reinforcement learning to learn with least data because data collection is expensive. in a real scenario.