Learning of Behavior Trees for Autonomous Agents
This addresses the challenge of domain-independent planning for autonomous agents, though it appears to be an incremental improvement combining existing techniques.
The paper tackles the problem of creating accurate system models for Automated Planners in real-world scenarios by proposing a model-free framework using Genetic Programming to derive optimal Behavior Trees for autonomous agents in unknown environments. Experimental results using the Mario AI benchmark show the framework can generate BTs that complete game levels at various difficulty levels.
Definition of an accurate system model for Automated Planner (AP) is often impractical, especially for real-world problems. Conversely, off-the-shelf planners fail to scale up and are domain dependent. These drawbacks are inherited from conventional transition systems such as Finite State Machines (FSMs) that describes the action-plan execution generated by the AP. On the other hand, Behavior Trees (BTs) represent a valid alternative to FSMs presenting many advantages in terms of modularity, reactiveness, scalability and domain-independence. In this paper, we propose a model-free AP framework using Genetic Programming (GP) to derive an optimal BT for an autonomous agent to achieve a given goal in unknown (but fully observable) environments. We illustrate the proposed framework using experiments conducted with an open source benchmark Mario AI for automated generation of BTs that can play the game character Mario to complete a certain level at various levels of difficulty to include enemies and obstacles.