I'm currently focusing on probabilistic trajectory planning by learning from demonstration in serial robots. I'm working on the problem of learning trajectories on Reimannian manifolds, especially SE(3), by means of its Lie algebra.
I'm also part of the HEAP project, working on the brenchmarking of different grasping algorithms on Franka Emika robot arms.
My previous research is mainly based on mechanisms theory and theoretical kinematics. I focused on the design of linkages and parallel manipulators that can change their operation modes and, in some cases, their permanent degrees of freedom. Similarly, I studied unexplored types of singularities in the configuration space of closed-loop linkages. These include cusp singularities, and tangential intersactions.
In most of this research, I used group and Lie theory, as well as more conventional geometry, like analysis of intersections of surfaces generated by linkages. Another tool that proved crucial for both analysis and synthesis is the higher-order kinematic analysis.