Novel method for preventing shin-collisions in six-legged robots by utilising a robot–terrain interference model

Solving the problem of interference and adaptive locomotion is crucial to the success of urban rescue missions that use legged robots. However, previous studies have mainly focused on locomotion over moderate terrain without considering leg interference. As a result, the current legged robots can easily get stuck when traversing rough terrain such as stairs. Therefore, a method for preventing interference is proposed so that legged robots can take full advantage of their inherent ability to walk over rough terrain. This study investigates the gait planning task for a six-legged robot climbing stairs. This paper proposes a novel robot–terrain interference model, which can be efficiently established online using a terrain map. The simplicity of the interference model allows the rapid computation of an interference criterion. Such a criterion assists in the foothold generation and configuration optimisation of the robot while traversing terrain irregularities. We performed simulations and experiments on a novel six-legged robot named QingZhui by utilising the online planning method realised with a hierarchical control framework. The results validated the effectiveness of the method.