Johan van Doornik

Papers

Total Citations

H-Index

About

Johan van Doornik is a robotics researcher whose work bridges the gap between human biomechanics and intelligent machine control. His primary research areas include human gait analysis, robotic control systems, and neural network applications in robotics. His most influential contribution is the development of a hydraulically actuated, 4-degree-of-freedom robotic platform for studying human gait under controlled perturbations. This apparatus, detailed in his 2007 paper (41 citations), allows researchers to apply precise velocity- or acceleration-controlled floor surface disturbances to freely walking subjects, providing a critical tool for understanding balance and locomotion. In the domain of robot control, van Doornik advanced the use of Radial Basis Function (RBF) neural networks. His 2011 work (11 citations) introduced an error-minimizing dead-zone in the learning dynamics, eliminating the typical requirement for Persistence of Excitation in desired trajectories—a practical improvement for real-world manipulator control. He further explored the robustness of such neural controllers by analyzing their stochastic stability under signal-dependent noise in the learning rule (2009). While his citation counts are modest, van Doornik's work is notable for its direct application to rehabilitation robotics and human-robot interaction, offering foundational insights into both experimental apparatus design and theoretically grounded adaptive control.