Humanoid Kinematics and Dynamics: Open Questions and Future Directions

Abstract

This chapter raises open questions about kinematics and dynamics for humanoid robots and sketches some possible future trends and directions. This requires to think about possible use cases in order to make good predictions about the future of kinematic and dynamic models. While initially bipedal walking with its conceptual challenges was of large concern, more recently applications such as service robotics, disaster mitigation, elderly care, and medical robotics become more important. The chapter focuses on three respective future directions and argues that, first, novel concepts in mechatronics have to embrace modularity and failure tolerance and to focus on durability and endurance, concepts that may lead to a larger parameter variance in the mechanisms. Second, it is expected that real-time simulation will play a larger role and may even be included in the control loops while coupling elastic, hydraulic, or pneumatic elements with the multibody dynamics. Third, the chapter makes the prediction that the hybrid combination of data-driven and classical modeling will be key toward becoming more flexible and possibly including more soft elements also in humanoid robots. The chapter concludes with the prediction that the future of kinematic and dynamic modeling is to combine advances in classical rigid-body computations with soft mechatronics, real-time simulation, and a mix of fast computing of classical algorithms and learning methods.