Obstacle avoidance control on omnidirectional vehicle robots using range sensor

There is a strong demand for intelligent vehicle robots in many fields. They are expected to assist in our daily lives. In order to implement them in practice, the robots should include useful functions such as path planning, navigation, and obstacle avoidance. This paper focuses on obstacle avoidance control techniques for multiple obstacles. The vehicle used as a platform is a mecanum wheel omnidirectional vehicle, which is driven on floors and includes a mounted scanner range sensor. The presented algorithm has the advantage of reducing multiple obstacle avoidance control to a single obstacle avoidance control scheme. Finally, experimental results are presented in order to evaluate the avoidance algorithms. © 2010 Wiley Periodicals, Inc. Electr Eng Jpn, 172(1): 58–68, 2010; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/eej.20933     

An automated weld seam tracking system for thick plate using cross mark structured light

This paper presents a weld seam tracking system using cross mark structured light. The hardware of the proposed system consists of a two degrees of freedom (DOF) welding robot, a camera with cross mark structured light, and two computers. The system has two parts namely visual sensing and motion control. In the visual sensing part, the cross mark of the structured light is utilized to set a region of interest (ROI). In the ROI, an adapted line fitting algorithm is employed to estimate the lines. Then, intersections of the lines are computed and used as the pin points for templates creating. During the matching process, a modified template matching is used to detect the edges of V-groove weld seam. By using this technique, a huge computational cost in image processing can be reduced, and therefore the tracking can be made in real time. The position based visual servoing with proportional-derivative(PD) and velocity feedback controller is designed for seam tracking. The experimental results show that the proposed method performs the real-time tracking efficiently with sufficient accuracy.