EvolvingBehavior: Towards Co-Creative Evolution of Behavior Trees for Game NPCs
This addresses the challenge of automating NPC behavior design for game developers, though it appears incremental as it builds on existing genetic programming methods for behavior trees.
The researchers tackled the problem of assisting game developers in creating non-player characters (NPCs) by introducing EvolvingBehavior, a tool for evolving behavior trees using genetic programming in Unreal Engine 4; in an evaluation within a 3D survival game, the tool produced behavior that approached designer goals compared to hand-crafted and randomly-grown trees.
To assist game developers in crafting game NPCs, we present EvolvingBehavior, a novel tool for genetic programming to evolve behavior trees in Unreal Engine 4. In an initial evaluation, we compare evolved behavior to hand-crafted trees designed by our researchers, and to randomly-grown trees, in a 3D survival game. We find that EvolvingBehavior is capable of producing behavior approaching the designer's goals in this context. Finally, we discuss implications and future avenues of exploration for co-creative game AI design tools, as well as challenges and difficulties in behavior tree evolution.