Petros Koumoutsakos

ETH Zurich, Massachusetts Institute of Technology, Harvard University

Papers

Total Citations

471

H-Index

About

Petros Koumoutsakos is a pioneer in computational science, merging fluid dynamics, machine learning, and optimization to solve complex problems in biology and engineering. His research focuses on understanding and harnessing the physics of flow, particularly in biological locomotion and collective behavior. A major contribution is his work on "Efficient collective swimming by harnessing vortices through deep reinforcement learning" (2018, 454 citations), which revealed how fish in schools exploit vortex wakes for energy savings, demonstrating that reinforcement learning can uncover optimal, nature-inspired swimming strategies. This work has profound implications for robotics and energy-efficient vehicle design. Koumoutsakos also advanced multi-objective optimization with studies like "Pareto Optimal Swimmers" (2017) and "Multi-objective optimization of artificial swimmers" (2017), systematically exploring trade-offs in swimming performance to guide bio-inspired design. His more recent "Learning efficient navigation in vortical flow fields" (2021) tackles autonomous navigation in complex currents, relevant for ocean robotics. Koumoutsakos’s impact is marked by high citation counts and his role in bridging computational methods with real-world applications, making him a leading figure in computational fluid dynamics and bio-inspired engineering.