MAR127 DIGITAL TWIN: Digital Twinning for Design and Operational Support

Summary of the project

Digital twins are digital representations of a physical object such as a ship, making use of numerical models and data. Digital twins are used to assist the maritime sector to develop new concepts, to design better ships, and to help the crew to detect anomalies, to plan maintenance, or to perform complex tasks on board. For the industry, the digital twin is a mean to increase competitiveness because it has the potential to bolster a faster, cheaper and safer project execution while optimizing the lifetime of their asset. In theory, the combination of the numerical models, large amount data measured by sensors on board and the current IT capabilities, makes the development and use of such digital twins possible.

However, it remains a challenge to make it happen in real world projects. This is because a digital twin requires several models and data to be integrated. In addition, it also needs to be synchronized, i.e. regularly updated to make sure it still describes the real ship accurately. How and when to integrate and synchronize the digital twin, while keeping it easy/flexible to use/maintain by various people is a difficult task. This project aims to explore the possibilities (and limitations) of digital twins. We will explore the challenges encountered when integrating and synchronizing the models and continuous measurements that form the digital twin base.

This requires a fast and efficient handling of the data, an effective communication with the users, but also robust and effective algorithms. The topic of digital twins has a tendency to become overly complicated and bogged down in complexity. To make the subject more tangible we will work with increments of complexity. As a first increment, we have set ourselves a challenge that we can all easily relate to and from which we fully understand the physical behaviour: a pendulum, representing for example a payload in a crane, or a rolling ship, will be made and coupled to its digital representation. Based on weather forecast, we will apply a force on the pendulum, and measure its motion. Using this relatively simple setup, we will forecast the motions, update the state of the pendulum when the motions are expected to exceed the maximum allowable threshold. We will also give people the possibility to change the state of the pendulum to test how robust the synchronization algorithms are, and to test how different algorithms can synchronize the real and the digital pendulum. Next increments in the project will be decided on together with the project partners, building on the experience gained from the first increment.

Goal of the project

This project is focused on the MMIP: Digital & Autonomous Shipping. It uses and contributes to key enabling technologies Digitization and AI / Data Science. Over the last years, we have gained knowledge on the model state and model parameter synchronization. Several components that are essential to the Digital Twin have also been investigated and a lot of knowledge was gained. This includes for example the (automatic) synchronization of numerical models, forecasting capabilities using neural networks, autonomous operations, anomaly detections algorithms. The objective of this work is to (1) connect the different (deep) knowledge components to each other; (2) show how a Digital Twin cycle actually works including applying the synchronizations; (3) identify the obstacles that need to be overcome to apply digital twin from concept to operations, for real-world problems. We will do this by developing a testbed coupled to a simulation environment. As a first increment, the case of a pendulum with which human can interfere, and being continuously measured at MARIN, is selected. The specific steps in the first part of this project are:

• Workshop with industry to define needs and expectations.
• Development of the first increment of the testbed and numerical environment.
• Development and iterative testing of standard connectors between physical and virtual environment related to integration & connectivity, synchronization, and human-machine interface
• Assess the performance (accuracy, robustness, speed, sensitivity…) of the digital twin and main obstacles to be solved towards real world application.

In the second part of the project, next steps will be defined based on the acquired experience and stakeholders input. As an example, the level of complexity may be increased by investigating the motions of the payload, in air or in water, with the crane mounted on a vessel in waves.

Motivation

In recent years, the interest of the industry on the use of so-called digital twins has grown significantly, together with the amount of data measured on board vessels. The digital twin is seen as a mean to increase competitiveness by helping the maritime sector to execute projects faster, cheaper and safer while optimizing the lifetime of their asset. However, the definition given to a digital twin, varies broadly as well as the application to solve real-world problems. The integration of multi-fidelity numerical models, of a large amount of measured data, and of the human being, from the concept to the operation phase is a challenge. The limited application of digital twins used from

concept to operations in the maritime industry, motivates this project to investigate how it can be used in our industry, and for what type of problems it should be used.

Innovativeness

It is proposed to build a digital twin which goes far beyond a simple numerical model because (1) it combines data, measurement techniques, systems and models in a modular manner, (2) it allows to switch between time scales, (3) it offers consistent multi-fidelity levels, (4) it offers interactive visualization to the observer, and (5) it accounts for the role of the human in the digital twin synchronization. Such a level of level of integration is required to be able to build the digital twin step by step, improve it project after project, to extend its use along different levels (descriptive, diagnostic, predictive/forecasting, prescriptive / decision-support, automated). The innovative approach followed in this project to use an experimental setup as a pilot case will not only allow to build and demonstrate the capabilities of the digital twins, but will also allow to identify the key obstacles that need to be overcome to solve the real world problems. The knowledge and insights learned in this project will be a step forward in the integration of Digital Twins in the Maritime industry.

Valorisation

Duration of the project

Start date: 01/0902024

End date: 31/03/2026