Learning NEAT Emergent Behaviors in Robot Swarms
Pranav Rajbhandari, Donald Sofge
TL;DR
The paper addresses how to learn policies that yield desired emergent behaviors in decentralized robot swarms. It extends NEAT to evolving neural networks controlling swarms, evaluating through full-episode fitness in CoppeliaSim with GT-MABs and Anki Vectors on area coverage and wall-climb tasks. Results show evolved policies can match or exceed hand-designed baselines, demonstrating robustness to sensing/actuation variations and highlighting the approach's applicability to complex swarm objectives. This work advances practical methods for automatic policy discovery in swarm robotics, potentially enabling scalable, task-specific emergent behaviors without extensive manual tuning.
Abstract
When researching robot swarms, many studies observe complex group behavior emerging from the individual agents' simple local actions. However, the task of learning an individual policy to produce a desired group behavior remains a challenging problem. We present a method of training distributed robotic swarm algorithms to produce emergent behavior. Inspired by the biological evolution of emergent behavior in animals, we use an evolutionary algorithm to train a population of individual behaviors to produce a desired group behavior. We perform experiments using simulations of the Georgia Tech Miniature Autonomous Blimps (GT-MABs) aerial robotics platforms conducted in the CoppeliaSim simulator. Additionally, we test on simulations of Anki Vector robots to display our algorithm's effectiveness on various modes of actuation. We evaluate our algorithm on various tasks where a somewhat complex group behavior is required for success. These tasks include an Area Coverage task and a Wall Climb task. We compare behaviors evolved using our algorithm against designed policies, which we create in order to exhibit the emergent behaviors we desire.