This paper describes the design and techniques of HAIBAO intelligent robot which is developed for Shanghai World Expo 2010. Compared with previous exhibition service robots, the HAIBAO robot has a more flexible motion ability, a more anthropomorphic interactive ability and a more intelligent cognitive and decision ability. Some key techniques including a four-wheeled omni-directional mechanism and its motion control and compensatory algorithm, the anthropomorphic interaction design, and the architecture of multi-tasks scheduling are introduced. During the Expo, totally 184 days, 37 HAIBAO robots have successfully serviced for the tourists by information providing, photography service, hall guiding, chat and various entertainments. Their robust, stability, flexibility and friendliness have been greatly commended.

The immanent existence of system latency greatly affects the control behavior of a closed-loop system. In order to reduce the influence induced by latency, this paper proposes a systematic method based on neural network to predict the motion of objects in a high-speed, dynamic, and competitive environment. We apply this method to the competition of RoboCup Small Size League, which greatly improves the performance of our control system.