Vijaykumar Gullapalli
University of Massachusetts Boston, University of Massachusetts Amherst, Princeton University
Papers
6
Total Citations
587
H-Index
5
About
Vijaykumar Gullapalli is a pioneering figure in the fields of reinforcement learning and robotics, best known for developing algorithms that enable machines to learn complex, real-world skills through trial and error. His seminal 1990 paper, "A stochastic reinforcement learning algorithm for learning real-valued functions," which has garnered 297 citations, introduced a groundbreaking method for continuous action spaces, laying the theoretical foundation for modern robot learning. Gullapalli’s work on skill acquisition, notably in his highly cited 1994 paper (215 citations), demonstrated how robots could master difficult tasks like peg-in-hole assembly and ball balancing using reinforcement learning, circumventing the need for explicit programming. He further advanced robotic dexterity by developing admittance learning for force-guided assembly (2002) and synergy-based control for redundant manipulators, enabling human-like hybrid position/force control. His research on shaping and direct reinforcement learning under uncertainty has been instrumental in making robots more adaptive and robust. With over 587 total citations across his key works, Gullapalli’s contributions have profoundly influenced both theoretical reinforcement learning and practical robotics, inspiring generations of researchers to build machines that learn from interaction.
Research Focus
Key Achievements
Top Papers
- 1A stochastic reinforcement learning algorithm for learning real-valued functions297 citations · 1990
- 2Acquiring robot skills via reinforcement learning215 citations · 1994
- 3Learning admittance mappings for force-guided assembly37 citations · 2002
- 4Skillful control under uncertainty via direct reinforcement learning23 citations · 1995
- 5
- 6Reinforcement Learning of Complex Behavior Through Shaping5 citations · 1997