Résumé : This article presents the design and real-time implementation of an optimal controller for precise steering control of a remotely operated underwater vehicle (ROV). A PI controller is investigated to achieve the desired steering performance. The gain parameters of the controller are tuned using the genetic algorithm (GA). The experimental response corresponding to the step waveform for the GA is obtained. A root-locus-tuned PI controller alongside a simulated-annealing-based PI controller (SAPI) is used to benchmark the response characteristics such as overshoot, peak time, and settling time. The experimental findings indicate that GAPI provides considerably better performance than SAPI and the root-locus-tuned controller.