A learning-based multisensor fusion approach for fine motion control of robots arms

We propose a concept for integrating multiple sensors in real-time robot control. To increase the controller robustness under diverse uncertainties, the robot systematically generates series of sensor data as robot state while memorising the corresponding motion parameters. From the collection of multisensor trajectories, statistical indices like principal components for each sensor type can be extracted. If the sensor data are preselected as output relevant, these principal components can be used very efficiently to approximately represent the original perception scenarios. After this dimension reduction procedure, a nonlinear fuzzy controller, e.g. a B-spline type, can be trained to map the subspace projection into the robot control parameters.