An improved and modified Probabilistic Roadmap (PRM) algorithm for non-circular holonomic mobile robot is proposed based on the consideration the shape and its kinematic constraint of non-circular holonomic mobile robot while moving from the initial point to the target point, especially to pass through the narrow passage. The path planning algorithm is also being modified to generate a shorter path distance in the shortest duration of time with the least number of turning points. The comparison of the conventional path planning algorithm will be carried out to determine the most optimal algorithm for modification to achieve the objectives. Then, the chosen path planning algorithm is PRM are improved by obstacle expansion and automatic selection for region of interest (AutoROI). Software simulation and analysis are performed on three different environment maps by a single holonomic mobile robot. The analysis of simulation revealed the performance of the modified path planning algorithm able to reduce elapsed time of PRM and increase the percentage of success path planning.