NAUTIS

Name: 
NAUTIS – Class-Compliant Digital Twin for Maritime Inspection and Targeted Robotic Waypoint Generation

The NAUTIS project develops an operationally proven Digital Twin that accurately represents a vessel’s true structural condition by fusing archival FEM/CAD data with historical Ultrasonic Thickness Measurement (UTM) records. It utilizes a hybrid Explainable AI (XAI) engine, strictly constrained by digitized IACS rules, to identify structural “hotspots” which are then translated into absolute 3D spatial waypoints. This shifts maritime robotic inspections from exhaustive, blind exploration to precise, predictive, and class-approved structural assessments, optimizing autonomous robotic exploration in GNSS-denied environments.

NAUTIS is a Class-Compliant Digital Twin designed to act as the analytical “brain” for maritime robotic inspections, addressing the critical disconnect between theoretical naval architecture and operational robotics. Currently, robotic platforms deployed in complex, GNSS-denied vessel compartments lack structural context, often scanning healthy and degraded steel indiscriminately. NAUTIS solves this inefficiency through three core technological pillars.First, NAUTIS executes advanced geometric decimation on computationally heavy CAD/FEM models to create a lightweight 3D structural mesh that preserves critical load-bearing topologies. It then projects theoretical stress concentrations and historical material loss (Ultrasonic Thickness Measurements) onto this mesh, generating a dynamic Corrosion and Stress Heatmap. Second, standard black-box machine learning is unsuitable for maritime regulatory compliance. NAUTIS innovates by digitizing the International Association of Classification Societies (IACS) Common Structural Rules into machine-readable Boolean logic. The hybrid architecture utilizes Supervised Learning to predict coupled stress-corrosion degradation but passes these probabilistic predictions through a deterministic Expert System (the IACS rules). Third, the system bridges structural engineering and robotics through dynamic matrix transformations, as it converts high-risk structural zones from the digital twin’s local coordinate system into absolute 3D spatial coordinates. These executable waypoints are pre-loaded onto the AUTOASSESS drone’s flight controller for autonomous offline inspection.

By supplying robotic platforms with mathematically prioritized risk-based 3D waypoints, NAUTIS minimizes indiscriminate scanning, and therefore reduces drone flight time and conserving battery payload. Also, by transforming raw post-flight sensor data into certified structural condition reports, NAUTIS delivers a reliable Decision Support System (DSS) for ship operators and classification societies.