One-shot path planning for multi-agent systems using fully convolutional neural network
This addresses the need for efficient, non-iterative path planning in robotics, particularly for multi-agent scenarios, though it is incremental as it builds on existing neural network methods.
The paper tackles the problem of path planning for multi-agent systems by proposing a fully convolutional neural network that generates complete paths in one shot, achieving optimal or close-to-optimal paths in over 98% of cases for single agents and 85.7% and 65.4% for two and three agents, respectively.
Path planning plays a crucial role in robot action execution, since a path or a motion trajectory for a particular action has to be defined first before the action can be executed. Most of the current approaches are iterative methods where the trajectory is generated iteratively by predicting the next state based on the current state. Moreover, in case of multi-agent systems, paths are planned for each agent separately. In contrast to that, we propose a novel method by utilising fully convolutional neural network, which allows generation of complete paths, even for more than one agent, in one-shot, i.e., with a single prediction step. We demonstrate that our method is able to successfully generate optimal or close to optimal paths in more than 98\% of the cases for single path predictions. Moreover, we show that although the network has never been trained on multi-path planning it is also able to generate optimal or close to optimal paths in 85.7\% and 65.4\% of the cases when generating two and three paths, respectively.