Papers
6
Total Citations
59
H-Index
3
About
Nat Dilokthanakul is a robotics researcher whose work sits at the intersection of human-robot interaction, adaptive locomotion, and intelligent control systems. His most impactful contribution is the development of a **visual goal human-robot communication framework using few-shot learning**, demonstrated in a robot waiter system. This work, which has garnered 34 citations, directly addresses the limitations of static-position control in dynamic environments like cafes and outdoor events, allowing robots to adapt to changing customer locations without pre-mapped goals. Dilokthanakul has also made significant strides in legged locomotion, proposing a **hybrid learning mechanism under a neural control network** for variable-speed quadruped walking (10 citations) and introducing **GRAB (GRAdient-Based Shape-Adaptive Locomotion Control)** for adaptive gait generation. His research further explores **risk-seeking exploration in deep reinforcement learning** (8 citations) and **dynamical state forcing on central pattern generators** for efficient robot control. Beyond these core contributions, he has contributed to the **factory of the future**, integrating drones and mobile manipulators for collaborative transport tasks. His work is notable for bridging bio-inspired control principles with modern machine learning, creating robots that are both more autonomous and more responsive to human needs.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Deep Reinforcement Learning with Risk-Seeking Exploration8 citations · 2018
- 4GRAB: GRAdient-Based Shape-Adaptive Locomotion Control3 citations · 2021
- 5
- 6Advanced Collaborative Robots for the Factory of the Future2 citations · 2021