In one of our recent posts, we introduced the Mesh Twin Learning (MTL) concept, which uses modern technologies to give industrial companies a competitive advantage by introducing micro-optimisations and sharing learned successes across the organisation.
MTL is pretty advanced concept, though. It requires data from IoT sensors, the creation of Digital Twins, training Machine Learning models and so on. All of this takes time and resources that not every organisation is prepared to spend upfront.However, aside from MTL, companies can still start benefiting from data-related technologies at even the early stages of their Industry 4.0 journey. So, let’s discuss these!
Digital Twins, IoT and Industry 4.0
Here, I want to elaborate on Industrial IoT and Digital Twins, which are the foundation of Industry 4.0.
Just to be clear, let’s have a quick recap:
- The IoT, or the Internet of Things, refers to devices with sensors and external connectivity. As such, these devices can record and submit data, opening up plenty of business options. In manufacturing and other industries, this covers everything from production equipment to logistics efforts.
- A Digital Twin is the virtual representation of a real-world object, process or system. The twin aggregates various types of data in relation to the object, from CAD designs and schematics to data gathered from IoT sensors themselves (both current and historical). Digital Twins of certain elements can also be linked together to create larger Twins of more complex systems. For instance, twins for individual machines can be connected to ‘twin’ an entire production line.
Digital Twins in Manufacturing
There are various ways Digital Twin can be implemented in context of manufacturing. However, we can broadly split them into three core categories:
- Digital Twins of individual products
- Digital Twins of production processes
- Digital Twins of performance
So, why are we talking about manufacturing specifically? Digital Twins are vital here because they allow for assessment and potential improvement without halting production. Previously, stopping a production line generated too much of a loss, so only essential maintenance was performed. Digital Twins give manufacturing companies advanced warning and notice, as well as the ability to fine-tune results in the virtual world first, before risking any changes in real processes.
With that out of the way, let’s explore the two key categories in more detail
Digital Twins of Individual Products
The Digital Twin of a product is created at the end of the respective product’s design stage. Once live, it is enriched with real-world data as the product is produced and utilised. As such, it provides the most up-to-date information on the current product’s properties as is possible without viewing physically handling the final product.
The most crucial key benefits of this are:
- Usage-Based Requirements – engineers can use real-world data concerning product usage and conditions when designing new versions of the product, which provides a better market fit and generates more values.
- Digital Product Traceability – Digital Twin technology provides a complete view on product information and enhances cross-functional collaboration. It improves the efficiency of change management in manufacturing and service processes. Additionally, these twins can be employed to provide manufacturing and service instructions as they are needed.
- Optimised Product Design – a digital model of a product, when enhanced with real-world performance data, enables complex simulations to improve the quality and performance of the newer versions.
One great example of such a Digital Twin for products comes from Whirlpool. The business combines real-world data from its smart, connected products with the product definition through a digital thread, and applies simulation capabilities through a Digital Twin to test prototypes with minimal investment.
Thanks to using Digital Twin technology, the organisation is able to minimise the time it takes to test new concepts – using facts instead of assumptions to drive ideation. This increases the speed of innovation and accelerates the introduction of new products to the field – two vital factors for success in a highly competitive market.
Digital Twins of Production Processes
Digital Twins of single assets, each a part of a larger production process, combine together to provide a whole new level of operational transparency and insight.
The digital thread of an operational workflow helps to connect the dots between disparate information systems. Some of the best use cases for these Digital Twins are:
- Connected Operations Intelligence – a Digital Twin for an entire production process combines data and information from disparate silos of assets and enterprise systems into a singular, real-time view of the whole process. Having such data and insights at hand increases operational performance and enables organisations to make better, informed decisions.
- The right information at the right time – when you connect the Digital Twin of a production process with Enterprise Resource Planning (ERP) and Manufacturing Execution Systems (MES), it is possible to provide adaptive work instructions in-context and increase operator productivity and improve production quality.
- Digitised Quality Assurance – Digital Twins provide data which can help to identify the source of quality issues and prevent them. Additionally, it is possible to translate the quality control expertise of in-house professionals to the digital system. Once it is in place, Machine Learning can be applied to the data, offering predictive quality alerts and their root causes.
A fantastic example of using Digital Twin technology in production can be found at Volvo. One of business’ goals is to improve flexibility and agility within manufacturing to accommodate modifications of configurations, according to their customers’ changing requirements. A digital thread, operation from design through to the entire manufacturing processes, was one of the biggest steps the company took to achieve engineering and manufacturing excellence.
Volvo also connected workers in the process to the digital thread and created a Digital Twin of quality assurance tasks in the plants. Combined, all of this resulted in:
- Improving operational effectiveness and cost savings, while getting closer to a 0 Part Per Million (PPM) quality / defective rate goal.
- Updating and validating engine configurations and QA checklist processes reduced these tasks from more than a day to less than an hour – freeing up time, money and other resources.
- The digitisation of the QA process is anticipated to save thousands of euros per workstation per year.
Digital Twins of Performance
A Digital Twin for operations collects the operational data from from various assets, such as the machines that form part of the production line, for example. Such a Digital Twin is fed with data from IoT sensors about all manner of parameters, such as temperature, pressure, vibrations, energy consumption and so on. When it comes to creating a complete view of the process – the more, the better!
Having such data in one centralised simulation gives businesses many advantages:
- Real-Time Performance Monitoring – when aggregating performance data, these types of Digital Twins can feed BI dashboards and provide accurate data visualisation for easy interpretation, helping companies to monitor state of the assets. Such combined data also makes it significantly easier (and quicker) to identify malfunctions. It also provides data for further analysis, which often leads to further process optimisation.
- Predictive Maintenance – data collected by a Digital Twin can be used to train Machine Learning models to monitor and optimise performance. Such ML applications can also detect machines with behaviour indicative of excessive wear and request required maintenance before the machine fails and causes downtime.
- Optimise Energy Consumption – having data about energy consumption and asset performance enables companies to identify key areas of improvement and implement energy efficiency measures, which can significantly reduce costs.
Gestamp, a Spanish automotive supplier, is a prime example of a company using Digital Twin technology to assess, refine and optimise performance. One of the most important goals for Gestamp is to reduce production costs as much as possible. This often means saving energy. The first step for the company was to implement solutions that collect energy consumption data from their production assets, such as presses and air compressors, in near real-time.
The created solution collects around 800 million data points every day. Armed with such data, the company was able to analyse the entire process and identify opportunities for improvements. Subsequent implementation of energy efficiency measures led to a total savings of 50 gigawatt-hours of energy in 2016 alone.
Combining planning with realistic digital projections has proved to be beneficial not only from a cost and time saving perspective, but also in terms of increasing revenue in the mid and long term. As we saw in this article, each application level of Digital Twin, be it on the product level, the process level or the performance level, brings with it its own advantages.
As a starting point, a single, non-critical asset can be copied in the digital space, where it can be experimented with at will. If this is successful, this technology can be applied to other individual assets, and processes too, until a stage of digital agility is reached, in which companies can foresee and implement changes at an ever-faster pace.
The ultimate level constitutes the connected factory mesh, where all insights collected on the micro and macro level are shared across production plants. The more Digital Twins are implemented, the more tangible this solution becomes.