Understanding How Component Failures Affect a Ship’s Capability
Dynamic Positioning systems rely on tight integration between sensors, controllers, thrusters, and power systems.
Image source: Breivik, Morten, Stig Kvaal, and Per Østby. “From Eureka to K-Pos: dynamic positioning as a highly successful and important marine control technology.” IFAC-PapersOnLine 48.16 (2015): 313–323.
Modern ships are complex systems that depend on many interconnected components to maintain safe and stable operation. For autonomous and highly automated vessels, it is not enough to detect when a component fails; it is equally important to understand how such failures or degradations influence the vessel’s overall capability.
This project addresses this challenge by representing the vessel as a causal and deterministic knowledge graph based on OPC UA information models. Each model describes a component’s function, dependencies and possible failure modes. Together they form a digital representation that can be updated in real time to show how failures or degradations affect total system capability.
Objective
The objective of this project is to develop modular OPC UA information models for key components of a Dynamic Positioning (DP) system and to explore how failures or degraded performance in these components influence the vessel’s overall capability. The combined model will act as a knowledge graph that links component level states to system level performance indicators and can serve as a foundation for future applications that use AI or analytical reasoning.
The student will define simple capability metrics, such as available thrust, power capacity, or position-holding accuracy, and implement logic that aggregates these from component level to system level. Concepts from, for example, Failure Mode and Effects Analysis (FMEA) will be used to identify relevant failure modes and to determine how each component failure influences the overall system capability. The result will be a simplified, real-time representation of the FMEA results where the effect of each failure is automatically reflected in the system-level performance during simulation.
(For an introduction to OPC UA, see What is OPC UA? (opc-router.com) or the OPC Foundation’s overview).
Work Description
- Develop modular OPC UA information models in Python, implementing an API interface that allows reading data, invoking methods and updating component states in real time
- Identify and describe failure modes for key components, such as thrusters, sensors, power systems, and control units
- Use FMEA results tomplement logic that propagates effects from component level to system level capability metrics such as thrust, energy or position accuracy
- Define and implement capability aggregation logic that computes total system performance indicators based on component states
- Connect the model to a DP simulator and visualize how capability evolves under different failure scenarios
Expected Outcome
A simplified digital representation of a DP vessel that updates its capability in real time according to component states. The work will provide a framework for analysing how failures propagate through interconnected systems and how system capability can be continuously evaluated and visualized.
Collaboration
This project is carried out in collaboration with Kongsberg Maritime, providing access to domain expertise, simulation environments, and insights into modern ship automation and Dynamic Positioning systems.
KM responsible: Sølve Raaen