Just what is a digital twin – and how can it help me?
The value of digitalization has been clearly demonstrated in many industries, and the marine industry is no exception. However, as the amount of data we accumulate has increased, the problem of how to use it effectively has become all the more acute. A “digital twin” – a virtual model of a physical asset that helps produce useful insights from raw data – is helping to solve this challenge.
The data quality challenge
However, a digital twin is only as good as the data you use to create the model. A wide range of factors can negatively affect the quality of data you use, including the type of equipment used, the placement of sensors, and the sensor settings. Although several methods – such as filtering, data normalization, and planned test runs – can improve the data we have for analysis, these all have intrinsic limitations and often leave too much room for interpretation. This uncertainty can lead to inconclusive results and the typical conclusion that yet more data is required; in the worst-case scenario faulty data can lead to bad decision-making.
Twinning is winning
To overcome these challenges, a digital or virtual twin of a physical asset like a vessel is created based on data collected from your asset and mimicking its desired characteristics. This maximizes the usefulness of collected data by modeling it for the specific asset in question, allowing you to build up a detailed understanding of its unique qualities. And best of all, with a digital twin one can simulate how the physical asset will behave and perform without needing to test it in the real world. This is as relevant for an individual piece of equipment as it is to a vessel, or even a whole fleet. You can also use the digital twin to optimize performance, for example predicting a vessel’s fuel consumption on different routes and with different speed profiles.
Increased data accuracy and reliability
A digital twin can also be made less susceptible to the imperfections of collected data and sensors by utilizing sensor-fusion technologies that combine different data sources to overcome the limitations of individual data feeds. For example, to model the performance of a ship more accurately, sensor-fusion technology can be used to combine ship data feeds with weather and current hindcasts, thus increasing the overall accuracy of the analysis. A sensor-fusion approach can also be used for estimating a vessel’s speed through water based on different data sources.
Learn from the past; predict the future
Digital twins help to provide answers as to why events happened. For example, if more fuel was burned on a voyage than expected, a simulation run using a digital twin might show there was bad weather, the ship was sailing against the current, or the trim or speed profile was not optimal. You can also rerun a simulation with different parameters to help you better optimize future voyages.
To simulate what might happen in the near future, users can combine a digital twin with multiple forecasts – for example, to see how a storm forecast en route might affect the ship’s performance, or the impact of a route change or engine profile change on the efficiency of the voyage. Digital twins also give a far more accurate picture for planning maintenance, maximizing the efficiency and reliability of equipment, and minimizing downtime.
Digital twins allow us to simulate and analyze what has happened in the past, optimize what is happening now, and predict what will happen in the near future – with far greater accuracy and reliability than previously possible. It’s no wonder that they’re increasingly being viewed as the future of data analysis.