Bio

Carlos Mastalli is a Research Associate in the University of Edinburgh. Before joining Edinburgh, he was a Postdoctoral Researcher in the Gepetto Team at LAAS-CNRS, Toulouse, France. He completed his PhD on “Planning and Execution of Dynamic Whole-Body Locomotion on Challenging Terrain” under the supervision of I. Havoutis, C. Semini and D.G. Caldwell. He was Invited Researcher in the Eidgenössische Technische Hochschule Zürich, Zurich, Switzerland, with J. Buchli in 2016.

The big picture of his research agenda is to find computational principles that describe the motor function, and at the same time be developed through robot experience during loco-manipulation tasks in robots with legs and arms. He is focused on building a new class of efficient algorithms for robot motor control with rich sensory data. His research is in the intersection of multiple disciplines: numerical optimization, model predictive control, whole-body control, inverse optimal control, and deep learning. He has contributions in those fields as well as successful open-source projects such as Crocoddyl.