In this work, a path planning strategy for both a car-like and a unicycle type assistive vehicles is presented. The assistive vehicles are confined to restricted environments. The path planning strategy uses the environment information to generate a kinematically plausible path to be followed by the vehicle. The environment information is provided by a SLAM (Simultaneous Localization and Mapping) algorithm implemented on the vehicles. The map generated by the SLAM algorithm compensates the lack of sensor at the back of the vehicles' chassis. A Monte Carlo-based technique is used to find the optimum path given the SLAM information. A visual and user-friendly interface enhances the user-vehicle communication allowing him/her to select a desired position and orientation (pose) that the vehicle should reach within the mapped environment. A trajectory controller drives the vehicle until it reaches a neighborhood of the desired pose. Several real-time experimental results within real environments are also shown herein.