Controllability-driven optimization of underwater ROV design

This paper presents a novel framework for optimizing the thruster configuration of Remotely Operated Vehicles (ROVs) using controllability metrics derived from linearized system dynamics. Central to the approach is the controllability Gramian, which quantifies how effectively inputs influence motion across six degrees of freedom. A cost function based on the Gramian’s eigenvalues enables systematic evaluation of thruster placements and orientations. Genetic algorithms and gradient descent are combined to explore both discrete and continuous design spaces. Validation is performed via a nonlinear 6-DOF MATLAB/Simulink model of the BlueROV2, a commercially available ROV with well-documented parameters. Results show that while optimized configurations improve theoretical controllability, practical factors like symmetry and force allocation must be addressed. The study demonstrates that controllability-driven optimization is a promising foundation for enhancing underwater vehicle performance.