MBSE-based development and verification of function-oriented ODDs for autonomous ships

Safe development and operation of autonomous ship functions depend on whether their operating conditions can be explicitly defined, quantified, verified, and monitored. This study proposes a model-based systems engineering (MBSE) method for developing and verifying function-oriented operational design domains (ODDs). Regulatory requirements, mission requirements, and operational risks are transformed into observable and decidable ODD conditions, with traceability established to functional architecture, physical deployment, verification scenarios, function-activation logic, and safe modes. The generic method comprises an ODD metamodel and an integrated design-time, test-time, and runtime verification process. Autonomous collision avoidance is used as a case study to instantiate the method. Its ODD is decomposed into environmental, positioning and perception reliability, communication, manoeuvring capability, collision-decision, and adaptive-safety conditions. These conditions are formalised as Object Constraint Language (OCL)-like constraints and Boolean ODD eligibility-gate (ODD_Gate) rules. Static model checking, deployment-consistency checking, integrated risk analysis, and representative scenario simulations are conducted to assess ODD formalisation, architecture mapping, function-activation control, and safe-mode switching. The results support the consistency of ODD-condition binding, deployment relationships, eligibility-gate evaluation, and safe-mode selection within the prescribed case-study inputs. The evidence is limited to model-level verification and does not establish the operational validity of the thresholds or collision-avoidance safety in real maritime traffic.