Implementing Ant-Gaiting-Based Biomimetic Locomotion for a Damaged Hexapod Robot

Abstract In this paper, we present biomimetic gait patterns for both undamaged and damaged hexapod robots, mimicking the locomotion of intact and amputated fire ants. We observed the gait patterns of intact and injured ants using a locomotion compensator and implemented a machine-learning image processing algorithm to accurately identify and track their movements. We modeled the locomotion using a Central Pattern Generator (CPG) with model parameters such as gait frequency, duty factor, and phase, leading to the derivation of one gait pattern for the 6-leg configuration and two major patterns for the 5-leg model. Furthermore, we designed and developed a biomimetic hexapod robot based on the dimensions of fire ants, controlling it using the derived motion patterns. The damaged hexapod robot demonstrated stable walking using these biomimetic gaits experimentally.