Biologically-inspired adaptive obstacle negotiation behavior of hexapod robots

Abstract

Neurobiological studies have shown that insects are able to adapt leg movements and posture for obstacle negotiation in changing environments. Moreover, the distance to an obstacle where an insect begins to climb is found to be a major parameter for successful obstacle negotiation. Inspired by these findings, we present an adaptive neural control mechanism for obstacle negotiation behavior in hexapod robots. It combines locomotion control, backbone joint control, local leg reflexes, and neural learning. While the first three components generate locomotion including walking and climbing, the neural learning mechanism allows the robot to adapt its behavior for obstacle negotiation with respect to changing conditions, e.g., variable obstacle heights and different walking gaits. By successfully learning the association of an early, predictive signal (conditioned stimulus, CS) and a late, reflex signal (unconditioned stimulus, UCS), both provided by ultrasonic sensors at the front of the robot, the robot can autonomously find an appropriate distance from an obstacle to initiate climbing. The adaptive neural control was developed and tested first on a physical robot simulation, and was then successfully transferred to a real hexapod robot, called AMOS II. The results show that the robot can efficiently negotiate obstacles with a height up to 85% of the robot's leg length in simulation and 75% in a real environment.

Learning by Association in Plants

“…In both experiments, the ability of seedlings to anticipate both the imminent arrival of light (“when”) and its direction (“where”) based on the presence and position of the fan indicates that plants are able to encode both temporal and spatial information and modify their behaviour under the control of environmental cues. This form of learning is ubiquitous in the animal kingdom1718, including all major vertebrate taxa and several invertebrate species19 and can also be implemented in artificial networks and machines20. Whilst the possibility that plants also learn by association has been considered by earlier studies2122, our current study provides the first unequivocal evidence.…”

The comparative investigation of the stick insect and cockroach models in the study of insect locomotion

“…Drawing inspiration from insects' compact and efficient designs and robustness coupled with plasticity, engineers have created successful models and machines (e.g. [39 , [76][77][78]). The present review has provided insights into interactions between sensory and motor control pathways in general, and highlighted concepts in adaptive motor control that may be useful in designing more responsive, robust, and flexible robots.…”

Decision Making Process of Hexapods in a Model of Complex Terrains

“…It of course follows that, alternatively, other nonlinear methods adapted for hexapod's decisions in complex terrain could be used, e.g. the tensor product models [31], methods for delivery vehicle routing problem [32], nonlinear multivariable systems using recurrent cerebellar models [33], neural control mechanisms [16], etc. Application of these methods to the problem of hexapod's decisions may find its place in the context of follow-up studies.…”

“…These methods are based i.e. on neural networks [15][16] or genetic algorithms [17]. Only a small number of methods is based on Reinforcement Learning (RL) [2,3,9,10], which is accepted as a method that describes the decision-making process of living organisms [18][19][20][21].…”