Dhaivat Dholakiya

Robert Bosch (China), Robert Bosch (India)

Papers

Total Citations

H-Index

About

Dhaivat Dholakiya is a robotics researcher whose work focuses on advancing quadrupedal locomotion through machine learning and bio-inspired design. His key research areas include deep reinforcement learning (D-RL), motion primitives, and active spinal mechanics for legged robots. Dholakiya’s major contributions center on developing novel control strategies that enable more efficient and dynamic walking in quadruped robots. In his most-cited work, "Trajectory based Deep Policy Search for Quadrupedal Walking" (8 citations), he introduced a trajectory-based policy search method using deep policy networks, which optimizes policies for entire walking cycles rather than individual time steps—a significant departure from conventional approaches. His paper "Realizing Learned Quadruped Locomotion Behaviors through Kinematic Motion Primitives" (6 citations) explores how minimal trajectory sets can be used to generate diverse walking patterns, drawing inspiration from biological motor control. Additionally, his study "Learning Active Spine Behaviors for Dynamic and Efficient Locomotion in Quadruped Robots" (4 citations) provides a simulation framework for investigating the role of spinal joint compliance and actuation in bounding performance, using the 16-DOF Stoch 2 robot. This work tackles the complex challenge of fast locomotion with active spines, offering insights that could lead to more agile and energy-efficient robots.