Investigation on the Influence of Obstacle Size in Path Planning by a Hybrid Model Combining an Improved A-star Algorithm and Digital Twin

The development of simulation technology has made it possible to create digital copies that are identical to a real system from geometry to dynamics. With reference to this, digital twin technology can help us determine the minimum collision-free distance between a robot and an obstacle on the virtual system, thereby planning the corresponding better path for the real system. In addition, the A-star algorithm was invented and many improvements were made to boost the efficiency of the original A* path planning algorithm. The advanced method shows the first breakthrough about the local path planning between a goal node and a current node, which has been already planned for the following search in the region of a current node. The local path will be adopted directly if it is safe and collisionless. This method also shows the second breakthrough about the application of the post-processing stage for path planning optimization by aligning the local path to lower a number of local paths along with path length. In this study, a combination of digital twin and an improved A-star algorithm was used for planning the robotic path in a light bulb assembly production line. The influence of obstacle size was also evaluated in terms of the efficiency of either method (i.e. A-star algorithm and digital twin) to further enhance robotic path planning when applied in practice to a system with obstacles of different sizes.