Designing Maritime Operations for new or complex situations traditionally relies on extensive consultation and full-scale trials, both of which rely on input from domain experts. These methods are often expensive, time consuming and have potentially uncertain outcomes. A method to discover high performing maritime operations is proposed by applying reinforcement learning techniques using scenarios implemented in commercial marine simulation technology. The approach is demonstrated with a simple case study using a short transit operation. Details and limitations of the method are presented and discussed.

This content is only available via PDF.
You can access this article if you purchase or spend a download.