Real-time Collision Avoidance for Robot Manipulators by Braking Trajectory Prediction
More Info
expand_more
expand_more
Abstract
The thesis contributes to the pre-collision strategies of robot manipulators by proposing a real-time implementable collision avoidance scheme. The collision is avoided by a braking controller that is activated by a braking trajectory predictor. The predictor provides distances of the end-effector to the object/human in the environment and braking is initiated when the predicted minimum distance to the object/human during braking is less than a certain threshold.