Adversarial Body Shape Search for Legged Robots

TL;DR

This paper introduces an evolutionary computation approach to generate adversarial body shapes for legged robots, revealing vulnerabilities in their design by disrupting walking performance through shape modifications.

Contribution

It presents a novel method combining evolutionary algorithms and deep reinforcement learning to identify adversarial body shapes that compromise robot walking stability.

Findings

Walker2d and Ant-v2 are more vulnerable to length attacks.

Humanoid-v2 is vulnerable to both length and thickness attacks.

Adversarial shapes often break symmetry or shift the center of gravity.

Abstract

We propose an evolutionary computation method for an adversarial attack on the length and thickness of parts of legged robots by deep reinforcement learning. This attack changes the robot body shape and interferes with walking-we call the attacked body as adversarial body shape. The evolutionary computation method searches adversarial body shape by minimizing the expected cumulative reward earned through walking simulation. To evaluate the effectiveness of the proposed method, we perform experiments with three-legged robots, Walker2d, Ant-v2, and Humanoid-v2 in OpenAI Gym. The experimental results reveal that Walker2d and Ant-v2 are more vulnerable to the attack on the length than the thickness of the body parts, whereas Humanoid-v2 is vulnerable to the attack on both of the length and thickness. We further identify that the adversarial body shapes break left-right symmetry or shift the center of gravity of the legged robots. Finding adversarial body shape can be used to proactively diagnose the vulnerability of legged robot walking.