In this paper, we discuss underwater walking robot technology to improve the quality of raw data in sector-scanning sonar images. We propose a strategy for an efficient and precise sector-scanning sonar image acquisition method for use in shallow, strong tidal water with a curved and sloped seabed environment. We verified the strategy by analyzing images acquired through a sea trial using the sector-scanning sonar installed on the CRABSTER (CR200). Before creating this strategy, an experiment was conducted to acquire the seabed image near a pier using a tripod and vertical pole. To overcome the problems and limitations revealed through image analysis, we established two technical strategies. In conclusion, we were able to achieve those technical strategies by using the CR200, which is resistant to strong current, and its six legs provide freedom of movement, allowing for a good sonar attitude.

The Crabster CR200 is an underwater walking robot inspired by crabs and lobsters, designed for precise seabed inspection and manipulation. It maintains stability and position on the seafloor, even in strong currents, by adjusting its posture through six legs, each with four degrees of freedom. The key advantage of the CR200 lies in its ability to resist drifting in strong currents by adapting its posture to maintain its position on the seafloor. However, information is still lacking on which specific posture generates the maximum downforce to ensure optimal stability in the presence of currents and the seabed. This study aims to determine the fluid forces acting on the CR200 in various postures using Computational Fluid Dynamics (CFD) and identify the posture that generates the maximum downforce. The posture is defined by two parameters: angle of attack and seafloor clearance, represented by the combination of the robot’s pitch angle and distance to the seabed. By varying these parameters, we identified the posture that produces the greatest downforce. Through a series of analyses, we identified two main fluid dynamic principles affecting the downforce on a robot close to the seabed. First, an optimal pitch angle exists that generates the maximum downward lift on the robot’s body. Secondly, there is an ideal distance from the seabed that produces maximum suction on the bottom surface, thereby creating a strong Venturi effect. Based on these principles, we determined the optimal robot posture to achieve maximum downforce in strong current conditions. The optimal underwater robot posture identified in this study could be applied to similar robots operating on the seafloor. Furthermore, the methodology adopted in this study for determining the optimal posture can serve as a reference for establishing operational postures for similar underwater robots.

Structural design under severe handling loads can be formulated as a constrained optimization problem in which structural mass is minimized while prescribed safety requirements are satisfied. In this study, a mathematical size-optimization model was developed for the frame of an underwater walking robot subjected to extreme launch and recovery loads. The frame geometry was parameterized using cross-sectional lengths and regional thicknesses as design variables. The optimization problem was formulated by minimizing the frame mass subject to an inequality constraint on the maximum equivalent stress and prescribed bounds on the design variables. Finite element analysis (FEA) was employed to evaluate the structural response associated with each design, and sequential quadratic programming (SQP) was used to iteratively solve the resulting nonlinear constrained optimization problem. The computational behavior of optimization was examined through the evolution of the design variables, objective function, and stress constraint. The optimization converged to a feasible design that satisfied the prescribed stress requirement while reducing the structural mass. The design evolution also showed that the geometric variables changed differently according to their locations in the frame, resulting in a redistribution of the structural response during the optimization process. These results demonstrate that the coupled FEA–SQP framework provides a systematic numerical approach for solving stress-constrained structural size-optimization problems and illustrates the application of mathematical optimization to the design of engineering structures subjected to severe loading conditions.