Papers
50
Total Citations
1,251
H-Index
15
About
Guillaume Sartoretti is a robotics and artificial intelligence researcher whose work spans multi-agent path finding, distributed reinforcement learning, and bio-inspired locomotion. He is perhaps best known for pioneering the PRIMAL framework — Pathfinding via Reinforcement and Imitation Multi-Agent Learning — which challenged the field's reliance on centralized planning by enabling decentralized, scalable coordination among large robot teams. With nearly 400 citations, PRIMAL stands as a landmark contribution to the multi-agent path finding (MAPF) community, subsequently extended in PRIMAL₂ to address lifelong, online variants of the problem and further refined through SCRIMP's attention-based communication mechanisms. Beyond path planning, Sartoretti has made significant contributions to collective construction robotics, autonomous exploration with deep reinforcement learning, and bio-inspired quadrupedal locomotion, investigating how body-leg coordination and central pattern generators can produce robust movement across unstructured terrain. His 2022 review of distributed reinforcement learning for robot teams underscores his role as a synthesizer and thought leader across the field. Collectively, his research advances a cohesive vision: intelligent, decentralized robot systems capable of operating effectively and autonomously in complex, real-world environments.
Research Focus
Key Achievements
Top Papers
- 1PRIMAL: Pathfinding via Reinforcement and Imitation Multi-Agent Learning399 citations · 2019
- 2
- 3
- 4
- 5
- 6
- 7Distributed Reinforcement Learning for Robot Teams: a Review34 citations · 2022
- 8
- 9Coordination of back bending and leg movements for quadrupedal locomotion32 citations · 2018
- 10Legged robots for object manipulation: A review31 citations · 2023