Robotic Relocalization Algorithm Assisted by Industrial Internet of Things and Artificial Intelligence

The Industrial Internet of Things (IIoT) scheme has the ability to integrate the computing center and the terminal actuator. It is usually the basis for the effective work of some IIoT systems that the terminal actuator can provide highly-precise location information. Under the assistance of the IIoT, we target at proposing a robot relocalization algorithm with high accuracy and stability in this paper. The relocalization method employing both semantic laser and landmark information is first designed, in which laser sensors are used to obtain quantitative information while semantic information is obtained using visual sensors. A pose derivation model based on the acquired landmark information is presented to correct the position of the actuator. In addition, the reinforcement learning is employed to dynamically select the optimal motion information during the relocalization process, based on which the positioning results are continuously optimized. The experimental results show that the proposed method has high accuracy and stability.