Design, swimming motion planning and implementation of a legged underwater robot (CALEB10: D.BeeBot) by biomimetic approach

In this paper, we propose swimming pattern generator (SPG) which is mimicking locomotion of diving beetles with a view point of bimomimetics for legged underwater robots. Firstly, the locomotion of the diving beetle has been observed and classified through experiments with a motion capture system consist of a high-speed camera and imaging processing software Image J. Subsequently, we analyzed coordinated patterns of rhythmic movements of the diving beetle's leg and formulated equations by employing Fourier least mean square fitting method corresponding to the obtained raw data from the motion capture system. Based on this, control parameters have been determined by understating their characteristics through the procedure of comparing produced swimming trajectories according to varying their values to observed motions of the diving beetle. For extended research of SPG (ESPG) to reduce the number of the parameters, they are expressed in a parameter space based on correlations of individual parameters and a newly configured parameter space is proposed through underwater experiments. Consequently the number of parameters has been successfully reduced for effective real time application of SPG. Furthermore, a six legged underwater robot has been developed by considering structural advantages that diving beetles have for their effective swimming. Underwater experiments have been conducted to confirm the feasibility of the proposed idea.