In this paper, we present a method that combines an improved dynamic window approach (IDWA) and an artificial potential field to avoid obstacles for autonomous wheeled robots with HyperOmni Vision. The experimental environment follows the rule of RoboSot avoidance challenge in the Federation of International Robot-soccer Association (FIRA) competition. There are three parts of this study, namely, (1) image processing, (2) an IDWA, and (3) an improved artificial potential field. In the avoidance challenge in FIRA, the robot must avoid eight obstacles that are placed randomly in the playing field. The results indicate that the threewheeled robot can avoid obstacles autonomously by using the method proposed in this paper. Our team won the FIRA avoidance challenge championship this year by the method proposed in this paper.

An autonomous robot is designed and implemented to support the service works. Many technologies were adapted, namely human-robot interaction order system, path planning and mapping in indoor, Robot Operating System (ROS), and Cyber Physical System (CPS). Many subsystems are developed and constructed based on ROS to implement a human-interaction action in completing service robot tasks. A human-robot interaction order system is designed for the convenient human servicing purpose. Therefore, the customer quickly finds the commodity information through the internet. This system recommends the commodity according to the content of the database. In addition, the customer makes an order by touch screen interface, and then it sends a command to control the robot to deliver the commodity. In the second subsystem, it includes path planning and mapping generation. The well-known simultaneous localization and mapping (SLAM) concept with a Rao-Blackwellized particle filter (RBPF) addresses the appropriated maps after completing the exploration of the indoor environment. In order to reach better robustness and agility, the hybrid path planning algorithm includes a global A* algorithm and local dynamic window approach (DWA) to navigation. In the third subsystem, the Gazebo makes the perfect construction of CPS design. Finally, the simulation and implementation of dual-arm robot navigation with the interactive order situation to efficiently support servicing actions of the robot and finish the required tasks.

No abstract