Go-Explore Complex 3D Game Environments for Automated Reachability Testing
This addresses the challenge of exhaustive level testing for AAA video games, which often ship with catastrophic bugs, by providing an automated solution that is incremental in applying an existing algorithm to a specific domain.
The paper tackles the problem of automated reachability testing in complex 3D game environments to find bugs like players falling through floors or getting stuck in walls, and shows that Go-Explore with simple heuristics outperforms baselines by covering a 1.5km x 1.5km game world within 10 hours on a single machine.
Modern AAA video games feature huge game levels and maps which are increasingly hard for level testers to cover exhaustively. As a result, games often ship with catastrophic bugs such as the player falling through the floor or being stuck in walls. We propose an approach specifically targeted at reachability bugs in simulated 3D environments based on the powerful exploration algorithm, Go-Explore, which saves unique checkpoints across the map and then identifies promising ones to explore from. We show that when coupled with simple heuristics derived from the game's navigation mesh, Go-Explore finds challenging bugs and comprehensively explores complex environments without the need for human demonstration or knowledge of the game dynamics. Go-Explore vastly outperforms more complicated baselines including reinforcement learning with intrinsic curiosity in both covering the navigation mesh and number of unique positions across the map discovered. Finally, due to our use of parallel agents, our algorithm can fully cover a vast 1.5km x 1.5km game world within 10 hours on a single machine making it extremely promising for continuous testing suites.