Innovative TPMS feet paired with deep reinforcement learning technology have been developed to enhance the energy efficiency of quadruped robots, reducing power consumption by up to 6.2%. Quadruped robots, widely used for tasks such as inspection, transportation, and rescue operations, face challenges with high energy consumption due to repeated leg movements compared to wheeled robots.
To mitigate this issue, Dr. Keun Park and Dr. Jung-Yup Kim from Seoul National University of Science and Technology in Korea have introduced porous TPMS robot feet that efficiently store and release impact energy. By combining these compliant feet with a deep reinforcement learning controller, the research team achieved a significant reduction in battery power consumption, particularly at walking speeds ranging from 0.4 to 1.0 meters per second.
TPMS metastructures, lightweight porous materials with excellent energy absorption properties, were utilized to design three 3D-printed foot models — primitive, gyroid, and diamond. The diamond design, with a relative density of 60%, was selected for its superior flexibility, energy absorption, and minimal energy loss. The integration of deep reinforcement learning allowed the robot to optimize its gait by leveraging the energy stored in the TPMS feet, thereby reducing motor workload.
Experimental results demonstrated that the optimized TPMS feet outperformed conventional solid feet by reducing battery power consumption by 1.4% to 6.2%. Moreover, the robot exhibited stable locomotion with minimized corrective movements, showcasing the efficacy of the controller in harnessing passive energy stored in the compliant feet for efficient quadruped locomotion.
Dr. Kim emphasized that this innovative approach could pave the way for the development of quieter and more energy-efficient quadruped robots suitable for various real-world applications such as indoor service, inspection, and logistics. The study detailing these advancements was published in the International Journal of Precision Engineering and Manufacturing-Green Technology on May 5, 2026.
Seoul National University of Science and Technology (SEOULTECH) spearheaded this groundbreaking research initiative. For more information about SEOULTECH, visit their website at https://en.seoultech.ac.kr/. For media inquiries, please contact Eunjung Son at 82-2-970-9166 or via email at [email protected].
About the Author
