Embedded Files
Ph.D. candidate at KTH Royal Institute of Technology
Former Master student at Delft University of Technology (Bio-robotics)
Former Visiting Student Researcher at Caltech (AMBERLab)
jorisv (at) kth (dot) se
I am interested in the planning and control of nonlinear and underactuated systems with a focus on robotics. When robots take over critical tasks, providing guarantees on completion, safety, and robustness will become ever more important.
I am a Ph.D. student at KTH Sweden under the supervision of Jana Tumova. My current focus is on temporal logic motion planning and control for multi-robot systems in space- and subsea environments.
Robust STL Control Synthesis under Maximal Disturbance Sets
Robust STL Control Synthesis under Maximal Disturbance Sets
This work focuses on synthesizing maximally disturbance robust controllers for a nonlinear dynamical system subject to an STL specifications. Existing notions of space- and time-robustness are model-invariant and cannot accomodate robustness where it is needed most. In this work we employ Hamilton-Jacobi reachability in order to find the maximum disturbance set such that the STL specification can still be satisfied.
Time-robust Multi-agent motion planning under STL
Time-robust Multi-agent motion planning under STL
This work focuses on generating temporally- or time-robust trajectories of multi-agent systems subjected to Signal Temporal Logic specifications. By using advantageous properties of BĂ©zier curves, we are able to do this efficiently. We provide theoretical guarantees of the soundness of our approach
preprint
Complex Dynamical Systems, walking like humans!
Complex Dynamical Systems, walking like humans!
Master thesis where reduced order model abstraction from humans subjected to expected and unexpected 'hole-in-the-ground' situations is used to describe the motion of Cassie. When Cassie is subjected to an unexpected down-step, it will adapt like a human and traverse these obstacles in a dynamically similar way!
Dynamic Walking 2022
Human-to-Robot Locomotion
Dynamic Walking 2022
Human-to-Robot Locomotion
Oral presentation at the Dynamic Walking conference at the University of Wisconsin on reflex-like compensation for expected and unexpected downstep scenarios.
Motion planning for drones in unstructured environments
Motion planning for drones in unstructured environments
Master project on quad-copter motion planning and control. Given the structured environment, a rough path is generated using RRT (Rapidly exploring Random Tree). The real-time control is performed by nonlinear MPC (Model Predictive Control) which considers unstructured objects
Page updated
Google Sites
Report abuse