Apparatus and method for controlling a robotic device
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ROBERT BOSCH GMBH
- Filing Date
- 2021-11-19
- Publication Date
- 2026-06-26
AI Technical Summary
Existing technologies struggle to effectively consider the relationship between the object and other objects, as well as the geometric conditions of the robot device, leading to a high failure rate in grasping and movement.
By training a deep image model, combining a hidden semi-Markov model and a neural network, robot trajectories and configurations are generated, taking into account the dynamic conditions of the object and the robot device, and optimizing the grasping and movement paths.
It significantly reduces the failure rate of robotic devices in grasping and moving objects, and increases the success rate.
Smart Images

Figure CN114516060B_ABST