S. Phani Teja

Indian Institute of Technology Hyderabad

Papers

Total Citations

H-Index

About

S. Phani Teja is a robotics researcher whose work bridges the critical gap between compliant mechanical design and intelligent control for mobile and humanoid robots. His primary research areas include modular robotics, compliant joint mechanisms, and multi-task reinforcement learning for complex robotic systems. Teja’s most impactful contribution is the development of a novel compliant modular robot designed for Urban Search and Rescue (USAR), with his seminal 2015 paper on stair climbing—his most cited work with 20 citations—demonstrating how passive joint compliance enables robots to ascend and descend stairs of varying dimensions without active joint actuators. He further refined this platform in a 2015 follow-up (5 citations) that specifically enhanced the robot’s ability to descend big obstacles through improved joint design. Teja also contributed to humanoid robotics with a 2016 paper (5 citations) detailing the design of an articulated-torso humanoid inspired by Poppy, modified with higher-torque MX-64 servos for increased load capacity and stability. Expanding into learning-based control, his 2018 work “DiGrad” (5 citations) introduced a multi-task reinforcement learning framework where shared actions and neural network parameters enable a single policy to perform multiple tasks concurrently—a significant step toward more efficient, general-purpose robotic intelligence.