Anatolii Borzilov

Papers

Total Citations

H-Index

About

Anatolii Borzilov is a leading researcher at the intersection of multi-agent reinforcement learning (MARL) and cooperative robotics, with a particular focus on multi-agent pathfinding (MAPF). His most influential work, the development of the POGEMA benchmark platform (2024), has already garnered 5 citations and addresses a critical gap in the field: the lack of scalable, partially observable environments for training and evaluating cooperative multi-agent systems. By designing POGEMA, Borzilov provides the research community with a standardized, challenging testbed that bridges the gap between idealized MARL simulations and real-world robotics constraints, such as limited sensor ranges and decentralized decision-making. This contribution is pivotal for advancing autonomous systems in logistics, warehouse automation, and search-and-rescue operations. Beyond this flagship work, Borzilov’s research consistently explores how deep reinforcement learning can be adapted to handle large numbers of agents operating under realistic communication and observability limits. His achievements are shaping the next generation of cooperative AI, making him a key figure for students and researchers interested in scalable multi-agent coordination and embodied intelligence.