Papers
35
Total Citations
307
H-Index
10
About
Diego Romeres is a robotics and machine learning researcher whose work sits at the intersection of data-driven modeling, robot manipulation, and intelligent planning. He has made significant contributions to Gaussian Process Regression (GPR) for learning robot dynamics, developing both black-box and semiparametric approaches that enable robots to identify inverse dynamics and navigate complex physical environments without relying on complete system models. His research on model-based reinforcement learning extends these methods to systems lacking velocity and acceleration measurements, broadening applicability to real-world robotic platforms. Romeres has also advanced the field of robotic manipulation, tackling challenging problems such as deformable object assembly, robust pivoting under frictional uncertainty, and anomaly detection during precision insertion tasks. More recently, he has embraced the integration of Large Language Models for interactive planning in partially observable environments and developed multi-level reasoning frameworks for autonomous assembly, reflecting a forward-looking research agenda bridging classical robotics with modern AI. His work has collectively garnered nearly 180 citations, with individual papers accumulating up to 26 citations, demonstrating meaningful community impact. Spanning simulation-to-real transfer, human-robot collaboration, and physics-informed learning, Romeres' research consistently addresses the gap between theoretical modeling and practical robotic deployment.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5Robust Pivoting: Exploiting Frictional Stability Using Bilevel Optimization20 citations · 2022
- 6
- 7
- 8
- 9
- 10