The DigiTOP Project offers opportunities to realise the potential impact that emerging digital technologies can have within manufacturing and maintenance settings. Whilst there is a whole host of emerging technologies, there are a number of uncertainties in terms of efficiency and effectiveness gains. Digital Twins (DTs) are one such technology that is emerging. This blog is focusing on the potential role that DTs can have.
A DT involves an integrated and living model that drives a business outcome by using continuous feed of data from multiple sources for a range of purposes. In the context of in-service support of high-value industrial assets, the core capabilities of a DT is the ability, to warn the user of system/physical errors, predict when faults will occur, optimise current procedures to obtain the lowest cycle time and lead time, forecast and adapt for changes to the object’s environment, obtain operational data (e.g. speed and temperature) in conjunction with the application of machine learning techniques, to cumulate damage and health assessments using thresholds as criteria to provide advance warning of failure. In addition, cognitive sensing can be used to improve the quality of the final product and enable improved designs. A combination of data analytics from multiple sources can improve the accuracy of the twin by filtering and comparing results from different tests. These can be compared to one another and provide results to the user, which will provide a better assessment in a complex interlinked real-world context.
The capabilities of a DT include the use of simulations to display historical data, which enables the live model to change over time, provide risk assessments with a scale that provides confidence in the results displayed and forecast revenue and cost. Similarly, machine learning, data analytics and AI are core components of a DT, which provides the system with the capability to monitor the health, identify the details and diagnose faults. Furthermore, additional components of a DT can include the use of a collection of data obtained from various sources, which are pulled together in one central location in the form of a dynamic model that shifts between design, build, operation and disposal. Thus, providing the user with enhanced decision-making capability throughout the life cycle of the product.
The capabilities of a DT can be used for the same product in different ways in order to adapt its use in different environments throughout the life cycle. Tao et al. (2017)¹ identifies three types of DTs:
- Parts twin – can focus on performance and anomaly detection, which is used for failure prediction (e.g. predicting the remaining life of a gas turbine).
- Process twin – can be used to schedule employees to maximise utilisation of resources (e.g. maximising the shop floor operations).
- System twin – can be used to optimise current processes against KPI’s (e.g. revenue, remaining product life and maintenance costs) to achieve the business’ requirements (e.g. organisation and final assembly operation)
In DigiTOP, across the project partners we will be focusing on different types of twins within the manufacturing and maintenance context. We will create an integrated DT that will enable to track the life cycle of a part, process or system. Lifecycle tracking enables the user to access data without the need for a physical presence to monitor the movements and changes. Furthermore, lifecycle tracking centralizes all paperwork associated with a product or person while onsite, which is useful when analysing behavior and fault patterns. In addition, lifecycle tracking can also be used to optimise production / maintenance methods to achieve targets such as cycle time, lead time, output rate and identify areas to reduce waste. The other benefits of our digital twin will include improved traceability, accuracy and faster access to information.
Dr. John Erkoyuncu
¹ Tao, Fei; Cheng, Jiangfeng; Qi, Qinglin; Zhang, Meng; Zhang, He; Sui, Fangyuan (March 2017). “Digital twin-driven product design, manufacturing and service with big data”. The International Journal of Advanced Manufacturing Technology. 94 (9–12): 3563–3576. doi:10.1007/s00170-017-0233-1.