Linear ship model promises simpler path to autonomous berthing

Despite decades of progress, fully autonomous ship berthing remains elusive due to the complex, non-linear dynamics of low-speed maneuvers. A new paper by Agnes N. Mwange, Taichi Kambara, Kouki Wakita, Kazuyoshi Hosogaya, and Atsuo Maki on arXiv (2607.01739) proposes a surprisingly simple workaround: model the ship as a time-invariant, continuous-time linear state-space system. The team estimated model parameters using the Covariance Matrix Adaptation Evolution Strategy (CMA-ES), a derivative-free optimization algorithm, applied to full-scale maneuvering data.

Validation results show strong agreement between the model's output and real-world measurements, demonstrating that a linear approximation can effectively capture low-speed ship motion. This finding challenges the conventional wisdom that non-linear models are necessary for accurate control. By proving that simplified linear models are viable, the research could significantly reduce the complexity of control systems needed for autonomous berthing. The next step involves testing the approach on different vessel types and under varying environmental conditions, but the initial results are a promising step toward making autonomous harbor operations a reality.