Megan Lecchi
Papers
1
Total Citations
15
H-Index
1
About
Dr. Megan Lecchi is a leading researcher at the intersection of reinforcement learning and dexterous robotic manipulation. Her work critically addresses one of robotics’ most pressing challenges: bridging the gap between high-performing simulation models and the constraints of real-world physical systems. Lecchi is best known for her landmark 2023 study, "Comparison of Model-Based and Model-Free Reinforcement Learning for Real-World Dexterous Robotic Manipulation Tasks," which has already garnered 15 citations and is shaping the field’s methodological direction. In this work, she systematically demonstrates that while Model-Free Reinforcement Learning (MFRL) excels in simulation, its prohibitive sample complexity and long training times render it impractical for complex real-world tasks. By contrast, she champions Model-Based approaches as a more sample-efficient and scalable alternative for physical robots. Lecchi’s contributions are pivotal for advancing practical, autonomous systems capable of learning intricate manipulation skills—from assembly to in-hand object reorientation—without requiring millions of simulated trials. Her research provides a clear roadmap for practitioners, making her a vital voice for students and engineers seeking to deploy reinforcement learning beyond the lab.
Research Focus
Key Achievements
Top Papers
- 1