Max Robotics — Global Service Robot Directory + US Certification

About

Holger Diedam is a leading figure in the field of optimal robot control, with a particular focus on real-time motion generation for humanoid robots. His research centers on the intersection of nonlinear model predictive control (MPC) and efficient numerical optimization, specifically through the development of fast direct multiple shooting algorithms. Diedam’s most significant contributions lie in enabling online walking motion generation without the need for predefined footsteps. His landmark 2010 paper, with 330 citations, demonstrates how MPC can be minimally modified to allow for automatic footstep placement, creating stable and adaptive gaits in real time. This work, along with his 2008 study on adaptive foot positioning via linear MPC (179 citations), has been foundational for the field of dynamic legged locomotion. His highly cited 2007 paper (337 citations) established the algorithmic framework for these advances, showcasing how direct multiple shooting can be optimized for real-time control. Through these contributions, Diedam has helped bridge the gap between theoretical optimal control and practical, reactive robot behavior, making him a key innovator in autonomous walking systems.