Thursday, February 23, 2016
11:00am 1170 TMCB
In recent years, indoor mobile robots are gaining more and more attention. At the same time, the requirement of robots' intelligence becomes higher. Especially for such tasks as simultaneous localization mapping(SLAM) and visual manipulation, there will be much computation with high complexity to acquire satisfactory result. However, the computation resource of mobile robot must consider the hardware cost and performance constraint, which cannot increase unboundedly. For this problem, cloud computing can provide effective solution with high cost-performance ratio. In this field, we tried to introduce cloud computing into SLAM, scene recognition and other visual tasks of indoor mobile robot. Parallel processing are also used to accelerate the real-time task. Depending upon ROS and Amazon EC2, we implemented low-cost robot system for indoor service with high performance. Experimental results show that cloud computing has bright future for indoor mobile robots.
Zhang Yunzhou received B.S. degree and M.S. degree in Mechanical and Electronic engineering from National University of Defense Technology, Changsha, China in 1997 and 2000, respectively. From 2000 to 2004, he served in China Airforce as a mechanical engineer. In 2004, he entered Northeastern University as a lecturer. He received Ph.D. degree in pattern recognition and intelligent system from Northeastern University, Shenyang, China, in 2009. Now, he is a professor with Northeastern University, China. His research interests include intelligent robot, image processing, and sensor networks. Under his instruction, more than 40 student teams have acquired national level award of science contest in robot, smart car and electronic system. He has published three books and over 50 articles in journals and conference.