Abstract

Abstract
Many people with disabilities find it difficult or even impossible to use traditional powered wheelchairs independently by manually controlling the devices. Intelligent wheelchairs are a very good solution to assist severely handicapped people who are unable to operate classical electrical wheelchair by themselves in their daily activities. This paper describes a development platform for intelligent wheelchairs called IntellWheels. The intelligent system developed may be added to commercial powered wheelchairs with minimal modifications in a very straightforward manner. The paper describes the concept and design of the platform and also the intelligent wheelchair prototype developed to validate the approach. Preliminary results concerning automatic movement of the IntellWheels prototype are also presented.

Abstract
Intelligent wheelchairs operating in dynamic environments need to sense its neighborhood and adapt the control signal, in real-time, to avoid collisions and protect the user. In this paper we propose a robust, real-time obstacle avoidance extension of the classic potential field methodology. Our algorithm is specially adapted to share the wheelchair's control with the user avoiding risky situations. This method relies on the idea of virtual forces, generated by the user command (attractive force) and by the objects detected on each ultrasonic sensor (repulsive forces), acting on the wheelchair. The resultant wheelchair's behavior is obtained by the sum of the attractive force and all the repulsive forces at a given position. Experimental results from drive tests in a cluttered office environment provided statistical evidence that the proposed algorithm is effective to reduce the number of collisions and still improve the user's safety perception. ©2010 IEEE.

Abstract
Many people with severe disabilities find it difficult or even impossible to use traditional powered wheelchairs independently by manually controlling these electrical devices. Intelligent wheelchairs are a very good solution to assist severely handicapped people who are unable to operate classical electrical wheelchair by themselves in their daily activities. This paper describes a development platform for intelligent wheelchairs called IntellWheels. The intelligent system developed may be added to commercial powered wheelchairs with minimal modifications in a very straightforward manner. The paper describes the concept and design of the platform, including the hardware and software, multimodal input interface and the intelligent wheelchair prototype developed to validate the approach. Preliminary results concerning automatic movement of the IntellWheels prototype are also described showing the autonomous movement capabilities of the prototype. © 2009 Springer-Verlag Berlin Heidelberg.

Abstract
Machine vision systems are widely used in assembly lines for providing sensing abilities to robots to allow them to handle dynamic environments. This paper presents a comparison of 3D sensors for evaluating which one is best suited for usage in a machine vision system for robotic fastening operations within an automotive assembly line. The perception system is necessary for taking into account the position uncertainty that arises from the vehicles being transported in an aerial conveyor. Three sensors with different working principles were compared, namely laser triangulation (SICK TriSpector1030), structured light with sequential stripe patterns (Photoneo PhoXi S) and structured light with infrared speckle pattern (Asus Xtion Pro Live). The accuracy of the sensors was measured by computing the root mean square error (RMSE) of the point cloud registrations between their scans and two types of reference point clouds, namely, CAD files and 3D sensor scans. Overall, the RMSE was lower when using sensor scans, with the SICK TriSpector1030 achieving the best results (0.25 mm +/- 0.03 mm), the Photoneo PhoXi S having the intermediate performance (0.49 mm +/- 0.14 mm) and the Asus Xtion Pro Live obtaining the higher RMSE (1.01 mm +/- 0.11 mm). Considering the use case requirements, the final machine vision system relied on the SICK TriSpector1030 sensor and was integrated with a collaborative robot, which was successfully deployed in an vehicle assembly line, achieving 94% success in 53,400 screwing operations.