What is a Digital Twin?
A Digital Twin is a virtual representation of a physical product or process that is used to understand and predict the performance characteristics of its physical equivalent. Digital Twins are used throughout the product lifecycle to simulate, predict and optimise the product and production system before investing in prototypes and physical assets. By incorporating multi-physics simulation, data analytics and machine learning capabilities, Digital Twins can demonstrate the impact of design changes, use cases, environmental conditions and countless other variables. Digital twins use data from sensors installed on physical objects to determine real-time object performance, operating conditions and changes over time. With this data, the digital twin evolves and is continuously updated to reflect any changes to its physical counterpart throughout the product lifecycle. This creates a closed feedback loop in a virtual environment that enables companies to continuously optimise their products, production and performance at minimal cost. The possible applications of a digital twin depend on the stage of the product lifecycle they model. In general, there are three types of digital twins: product, production and performance.
The combination and integration of the three digital twins as they evolve together is known as a digital lattice. The term “lattice” is used because it is intertwined in all stages of product lifecycles and production systems and brings together data from all these stages.
Digital twins of a PRODUCT can be used to virtually validate product performance, and also show how products actually perform in the physical world. This “digital twin of a product” provides a connection between the virtual and physical world, allowing the performance of a product under various conditions to be analysed and adjustments to be made in the virtual world to ensure that the next physical product performs exactly as planned in the field. It doesn’t matter if the systems and materials are complex: digital product twins help navigate this complexity to make the best possible decisions. This avoids the need for multiple prototypes, reduces overall development time, improves the quality of the final manufactured product and enables faster interactions in response to customer feedback. The cycle time for product commissioning is usually almost linear, from idea and design, to mechanical manufacture, to electrical installation, and then programming. By contrast, with the use of the digital twin we can overlap different stages of execution. It is possible to start programming the line or machine before it is manufactured or during its design to verify correct operation. This reduces the time by approximately 30% and could reduce prototyping costs if any.
A PRODUCTION digital twin can help validate how efficiently a manufacturing process will work on the shop floor before anything actually goes into production. By simulating the process with a digital twin and analysing why things happen using the digital lattice, companies can create a production methodology that maintains its efficiency under various conditions. Production can be further optimised by creating digital product and production twins, companies can avoid costly equipment downtime and even predict when preventive maintenance will be necessary. Smart products and plants generate massive amounts of data on their utilisation and efficiency.
The PERFORMANCE digital twin captures this data from running products and plants and analyses it to provide information for critical decision-making. Thanks to digital performance twins, companies can:
- Generate new business opportunities
Gain insights to improve virtual models
Capture, aggregate and analyse operational data
Improve product and production system efficiency
Regardless of the above classification, the digital twin can be categorised according to its scope of use in relation to the product cycle. - Pre-sales to analyse a solution before selling it, planning and design. A model can be created to verify that the designs are correct for the scope of the project. It is also feasible to emulate the control software (PLC, robots, etc.) to validate it.
After-sales. To correct incidences in the software, improve the performance of the line or analyse possible changes in the lines without the need to implement it physically.
Problems minimised with the use of the digital twin
The sale of lines has a number of problems that the digital twin model can solve without the need to physically replicate it on site. The main ones are:
Size of the lines to be installed. It is not feasible to install the lines in the manufacturer’s facilities due to their dimensions and therefore it would not be possible to test their control and efficiency.
Absence of intermediate machines. On certain occasions, machines are purchased from third party suppliers and/or provided directly by the end customer, which means that they cannot be integrated into the line for full testing.
Inability to reproduce production speeds. Not enough product is available or it is unfeasible to introduce products by hand. This results in machinery being installed at the customer’s premises without being fully tested, resulting in: long start-up times due to the need for more time to debug control programs. Lost time in replacing parts that do not work as intended.
Qualified or experienced personnel to start up the machinery.
State of the technology
Digital Twin technology is evolving and being adopted by different sectors of the global economy and the industrial sector is no different. The concept of the Digital Twin is increasingly being heard and this is driven by the rise and settlement of Industry 4.0 and that the technology is becoming more mature and allows for more sophisticated methods. Morgan Stanley indicates that the market for cybersecurity, a prerequisite for the connected industrial, will reach $183 billion by 2020. Gartner gives even higher figures for the IoT industry. It predicts that half of large industrial companies will have adopted the digital twin model providing them with 10% efficiency gains. In other words, industry is more connected than ever before, and in an increasingly secure way. IoT projects are more scalable and are starting to take off.
There has also been strong investment in automation, robotics, sensorisation…. However, much of these efforts have been directed at solving specific problems. Now the challenge is to connect all the efforts to maximise their benefit, and that is where digital twins can come into play. According to a study by Moning Consult and IBM, Spain is a pioneer in Europe in the adoption of artificial intelligence (AI) technology, where the concept of the digital model has a place, in which 82% of Spanish companies are said to be implementing or exploring the incorporation of artificial intelligence technologies into their processes, some 9 points above the average of the large European countries (Germany, France, Italy, United Kingdom and Spain). Implementation is focusing on the areas of security (42%), process automation (31%) and customer service (29%).
Spanish companies indicate that the main barrier to further progress is the lack of expert knowledge (39%), highlighting the training and adaptation challenge that the development of artificial intelligence represents for companies and society as a whole. Despite this barrier, the vast majority of Spanish companies (72%) say that trust in technology is the factor that is contributing most to creating a culture favourable to the adoption of AI in their organisations.