Legged robot locomotion on natural soft terrain

• Applicant: Professor Dr.-Ing. Raul Fuentes
• Subject Area: Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 550114247

In this project we will demonstrate, quantify, and characterise how using an accurate geomechanics modelling of leg. Achieving this will entail scientific developments in both robotics and geomechanics in a highly multidisciplinary project that could transform how legged robots (and others) are designed to operate in unstable terrains-ground interaction improves robot’s general efficiency. We hypothesize that the performance and reliability of legged robot locomotion on soft substrates can be improved by replacing currently used, simpler ground models used by roboticists (springer/damper, friction cone), by more precise geomechanical models. Under this premise, providing a better estimate of ground reaction forces and foot displacements through geomechanical models of soft substrates, robot control can be improved, leading to a wider viable control parameter space and a better robot performance, including a higher robustness and more energy efficient legged locomotion. We hypothesize that walking on soft substrates requires locomotion control strategies and parameters individually set for substrate types and substrate parameters like water content and pre-compression state. Until now, no geomechanics models were utilized to control the locomotion of legged robots in real-time. We propose that, in a stepwise process, geomechanical models work in combination with MPCs and simplified robot models and controllers (LIPM, TSLIP, ZMP), remaining real-time capable.

• DFG Programme: Research Grants
• International Connection: Belgium
• Partner Organisation: Fonds Wetenschappelijk Onderzoek - Vlaanderen ; Research Foundation Flanders (FWO)
• Cooperation Partner: Professor Dr. Alexander Badri-Spröwitz