Adaptive Locomotion of Soft Microrobots
Networked Control and Communication
Controller Learning using Bayesian Optimization
Event-based Wireless Control of Cyber-physical Systems
Model-based Reinforcement Learning for PID Control
Learning Probabilistic Dynamics Models
Gaussian Filtering as Variational Inference
Autonomous Robotic Manipulation
We have developed algorithms which enable an autonomous manipulation system to grasp a wide range of objects and to perform a certain number of manipulation tasks, such as drilling, using a stapler, unlocking a door with a key or changing a tire []. More generally, we are interested in providing complete integrated systems with a certain level of autonomy. The motion system of the architecture is centered around two key concepts: 1) the control of contact interaction through high performance force and impedance control and 2) the use of optimization to generate reproducible motions of high quality.
Unlike an industrial robot, such a system has to adapt to situations it has not previously encountered. Therefore, it needs to infer properties of its surroundings using sensors, learn from experience and be robust to disturbances. In the context of this project we develop algorithms which address these challenges. A few examples are:
- Data-driven grasping algorithms []
- Learning approaches for force control []
- Algorithms for reactive movement adaptation in face of disturbances [
] - 3D object tracking algorithms using a range camera []
Members
Publications