Digital Twin of Autonomous Surface Vessels for Safe Maritime Navigation Enabled through Predictive Modeling and Reinforcement Learning

TL;DR

The paper addresses safe autonomous maritime navigation by extending a Unity-based digital twin (DT) of an autonomous surface vessel (ASV) with predictive target tracking using AIS and LiDAR, and a nonlinear model predictive control–based predictive safety filter (PSF) to guarantee safe autonomous operation. The approach integrates predictive tracking (ellipse fitting and sensor fusion) with a safety-critical NMPC framework, enabling what-if scenario analysis and prescriptive control within the DT. Key contributions include (i) a predictive target-tracking pipeline using AIS–LiDAR fusion and numerically stable ellipse fitting, (ii) a PSF that enforces safety constraints and terminal invariance via a semidefinite-programmed ellipsoidal set, and (iii) demonstration in a 3D Unity DT showing improvements in collision avoidance and path following for RL-driven control, with real-world AIS data enhancing SITAW. The results indicate that the DT can provide predictive, prescriptive, and near-autonomous decision support for safer maritime operations, while also highlighting integration-specific challenges in near-shore scenarios and the need for broader physical modeling and sensing in future work.

Abstract

Autonomous surface vessels (ASVs) play an increasingly important role in the safety and sustainability of open sea operations. Since most maritime accidents are related to human failure, intelligent algorithms for autonomous collision avoidance and path following can drastically reduce the risk in the maritime sector. A DT is a virtual representative of a real physical system and can enhance the situational awareness (SITAW) of such an ASV to generate optimal decisions. This work builds on an existing DT framework for ASVs and demonstrates foundations for enabling predictive, prescriptive, and autonomous capabilities. In this context, sophisticated target tracking approaches are crucial for estimating and predicting the position and motion of other dynamic objects. The applied tracking method is enabled by real-time automatic identification system (AIS) data and synthetic light detection and ranging (Lidar) measurements. To guarantee safety during autonomous operations, we applied a predictive safety filter, based on the concept of nonlinear model predictive control (NMPC). The approaches are implemented into a DT built with the Unity game engine. As a result, this work demonstrates the potential of a DT capable of making predictions, playing through various what-if scenarios, and providing optimal control decisions according to its enhanced SITAW.

Figures (12)

• Figure 1: Capability scale of digital twins
• Figure 2: Implementation of a standalone digital twin in the Unity game engine.
• Figure 3: Implementation of a descriptive digital twin in the Unity game engine.
• Figure 4: Implementation of a diagnostic digital twin in the Unity game engine.
• Figure 5: Sensor fusion principle. AIS and LiDAR measurements are separately propagated through Kalman filters using a constant velocity model.