Powered wheelchairs have benefited millions of people with moving difficulty around the world. However, severed disabled people may need smart wheelchairs with the ability similar to an autonomous mobile robot such as obstacle avoidance, travelling through a narrow passage, localizations, following targets… This paper describes the implementation of a localization algorithm on a semi-autonomous wheelchair platform developed for educational purposes at Ho Chi Minh City University of Technology and Education. The algorithm uses data from a laser scanner SICK LMS 291. The Extended Kalman Filter is used to better estimate the position of the wheelchair. D* search is used to plan the path for the wheelchair to move from its initial position to the goal position. The result shows the wheelchair could follow a predefined path very well thanks to the ability to localize itself.
