This project is aimed at developing an intelligent autonomous mover (iAMEC in short) for environments surrounding with crowds. The main applications include automatic customer service for shopping centers. iAMEC can avoid the obstacle while navigating to the target location. It can also classify the pedestrians from other objects correctly and analyzes their behaviors for danger assessment. In addition, an indoor location system and a traffic control center are established for iAMEC. The indoor location system reports the location of iAMEC constantly to the traffic control center, which is in charge of the dispatch and route planning of iAMECs. The key technologies developed include (1) obstacle detection, (2) AI/deep learning based obstacle classification, (3) behavior predictions, (4) AI edge inference engine, and (5) traffic control and navigation. Preliminary results have been achieved in mover design, image based object detection & classification, ultrasound array radar, CNN hardware accelerator chip, cloud server, mover navigation control and optimal route planning. A simplified system is also set up for the demo purpose.

Autonomous Mover, Obstacle Detection, AI/Deep Learning