Intention modeling and inference for autonomous collision avoidance at sea

Abstract

The open wording of the traffic rules of the sea, COLREGS, and the existence of unwritten rules, make it essential for an autonomous ship to understand the intentions of other ships. This article uses a dynamic Bayesian network (DBN) to model and infer the intentions of other ships in open waters based on their observed real-time behavior. Multiple intention nodes are included to describe the different ways a ship can interpret and conflict with the behavioral rules outlined in COLREGS. The prior probability distributions of the intention nodes are adapted to the current situation based on observable characteristics such as location and relative ship size. The resulting model is able to identify situations that are prone to cause misunderstandings and infer the state of multiple intention variables that describe how the ship is likely to behave. Different collision avoidance algorithms can use the resulting intention information to better know if, when, and how to act.

Intention modeling and inference for autonomous collision avoidance at sea