| Definition | VoxelGym2D is a grid-based reinforcement learning environment for efficient path planning using voxel representations. |
| No of environments | 1 |
VoxelGym2D is a reinforcement learning environment designed for grid-based path planning tasks, utilizing a voxel representation to simulate spatial challenges. This environment allows agents to navigate through structured grids, where they learn to identify and reach designated goals while optimizing their paths. Through training with algorithms like Proximal Policy Optimization (PPO), agents demonstrate improved performance, as indicated by increasing episode rewards and decreasing episode lengths, showcasing their ability to efficiently solve navigation tasks.
| Example observations | |
| Example actions | |
| Example reward |