In recent years there has been a great deal of buzz around the concept of Digital Twins. According to IDC’s 2019 Digital Transformation FutureScape, 30% of G2000 companies will have implemented ‘advanced digital twins’ to optimise their operations by 2020, the vast majority of which will be industrial enterprises. While discussions of digital twin have framed this technology as a modern business imperative, the first mention can be dated back to NASA in the 60s with the Apollo 13 mission. For enterprises deep in multitudes of digital transformation initiatives, digital twins can be convenient to lump into the category of nice to haves to add to the end of the to-do list. The truth is, digital twin technology is available today – and if you’re an industrial enterprise, it’s very likely that your existing technology can start compounding additional value unified through a digital twin strategy. In doing so, enterprises can expect to achieve outcomes like differentiation for their products, improved operational effectiveness across processes, and optimised productivity for their people.
Digital twins are digital models that virtually represent their physical counterparts. This virtual representation of a physical product, an operational process, or a person’s task is used to understand or predict the physical counterpart by leveraging both the business system data that defines it and its physical world experience captured through sensors.
What started out with the capture and storage of enterprise data has slowly crept out of the server closet. Sensors are everywhere, and telemetry data is being created by not only Smart, Connected Products, but for entire enterprise processes and systems, and with augmented reality and other people-centered technologies being adopted, there are even digital connections to people. The result is that today this data exists in siloes across organisations, each possessing limited context which in turn limits its value. Enterprises have recognised the need to integrate this new data but have been struggling with the strategy to do so in a scalable and effective way. Systems integration spending is growing rapidly year over year, and accounts for increasingly large portions of industrial companies’ spend. One key objective of this spending is to map data from currently isolated technology siloes as companies recognise the value of centralising and simplifying the process of information discovery and analysis. In addition to providing universal data access, enterprises also recognise that exploring the relationship between interrelated data sets offers higher order insights than any one set alone. This data unification is a necessary pre-requisite to building a digital twin, and is often referred to as a digital thread.
With new technologies like augmented reality and IoT creating and demanding vast amounts of data from disparate sources, it is more important than ever that companies have a clear strategy as to how and why they will integrate their various operational and information technology. The digital thread is the essential groundwork that can turn the common dilemma of spiraling cost and complexity of digital transformation into an opportunity to enable faster time to value, greater agility in change management, and more datadriven decision making. A digital twin is a model for contextualising, analysing, and realising the value of the digital thread in a way that enables it to be acted upon and scaled across multiple solutions and front-end applications. In other words, the value of a digital thread is manifest through the discrete use cases served by digital twins.
The time is now for industrial enterprises to build out their digital twin strategies. With the maturity of the enabling technologies and digital thread initiatives reaching critical mass, many companies are taking stock of their current capabilities and moving quickly to fill the gaps. The way digital twins are delivered through various lenses can vary greatly based on the specific use case being pursued, but core considerations should be addressed based on necessary capabilities. Companies seeking to advance their digital twin strategy will benefit from organising current and future capabilities into the following framework. From there, specific use cases can be plotted to organise requirements and develop a plan of action.
A digital twin requires you to combine the digital definition from related business systems with real-world data and insights from the physical world via sensors. Companies must decide what source data they will include, for example manufacturing process plans or operating procedures that define a process combined with the real-world telemetry and sensor data from manufacturing and production environments. Adding additional sources to its definition, for example supply chain data from an MES system, can drive increased overall context for the twin as well as unlock additional use cases without rework. Additional technologies continue to add to potential sources of insight. In the future, with the bounty of sensor data coming online through AR devices, people and the spaces they inhabit (factories, buildings, etc.) will be defined and integrated into twins as well.
Digital twins give unified insight into the data connected by the digital thread. Once a use case is understood, unique identification and organisation of data surrounding an individual product, process, or person can be mapped and organised to inform the twin model. It could be contextualised into an overall process, enhanced with behavioral data, or used to align to desired KPIs. Understanding the over-arching goals of the twin will help to contextualise it into the type of digital twin model that makes the most sense.
Analytics can be used to add value for certain use cases to inform business decisions with greater accuracy and unlock hidden insights, or the value can be self-evident. Analytics could be applied on the mapped data to answer questions along common frameworks, for example descriptive answers to questions like how a product is performing, to diagnostic questions around cause of failure, and predictive or prescriptive questions that simulate potential scenarios and optimise performance outcomes.
Orchestration is where these insights are put to task. Triggers can be created that automate or direct actions based on the result of the answer or analysis. For example, a process trigger could be put into place to dispatch a technician or create a customer service ticket for a product failure. You could automatically propose remote configuration updates based on performance characteristics. It is also possible to deliver updated KPIs and worker priorities based on customer or supply chain activities. Whatever questions you seek to answer, a corresponding action can be orchestrated to react, and measurable KPIs and outcomes can be captured.
Digital twins are delivered or interacted with through a frontend UI or ‘lens’ that is role or task-based and specific to a given use case. They can be delivered through interfaces such as desktop and mobile devices, and emerging technologies like augmented reality are provide additional options. In fact, augmented reality provides the capabilities to capture spatial data related to environments and workers to eventually develop digital twins of these previously undefined spaces and processes. These technologies also offer enhancements to the fidelity and user experience of digital twins and make use cases accessible to new stakeholders such as deskless workers.
Contact InVMA to learn more about how our network of strategists and partners as well as our unique technology portfolio can help you unlock the value created at the convergence of the physical and digital worlds through the use of digital twins
Design review and demonstration have become increasingly important for product development and sales. However, there are several obstacles that can prevent a company from fully realizing their product’s potential.
ABI Research explores how Augmented Reality improves product development and sales in three key areas: