Digital twin: a popular term with many definitions. Not everything that goes through life under this heading is really a Digital Twin according to Cognizant. We talk to two experts about what exactly it takes to build a true digital twin, what such a thing is for, and what the added value can be for an enterprise.
Everyone today talks about digital twins, but what exactly does that technology mean? “The term is used enthusiastically in different sectors,” agrees Christophe Stas, Associate Director of Life Sciences Manufacturing at Cognizant. “But everyone understands it differently.” Time, then, for a definition. Cognizant explains the meaning of digital twins within Life sciences, highlights their real value to an organization and identifies the critical factors for successful implementation.
Digital twins in the life sciences are virtual models connected to their physical counterparts that mimic the behavior of biological processes such as cell culture growth in a bioreactor. These twins use real-time data architectures to feed the models with sensor data, predict process development and suggest optimal control actions to maximize production yield. In the most advanced implementation, a digital twin can directly control the physical process fully automatically.
Connection to the real world
Dr. Elisa Canzani, Data Science Lead at Cognizant, clarifies, “A digital twin is a model that relies on sensors to provide (near) real-time feedback or updates on its physical counterpart. Many organizations confuse a digital model with a digital twin. We see the digital twin as an end-to-end connected reality that leverages data streams to provide a dynamic representation of physical processes in real time.”
A digital twin is a model that relies on sensors to provide real-time feedback or updates on its physical counterpart
Dr. Elisa Canzani, Data Science Lead Cognizant
That distinction is important. The Digital Twin only really comes to life when it is connected to the physical manufacturing process. Digital twins must be connected to the physical sensors and process controller to use actual production data to feed the virtual models, according to Cognizant. In turn, the digital twin can predict and optimize production parameters to improve the efficiency of the manufacturing process through an open or closed feedback loop.
From process to platform to model and back
Canzani clarifies what such a system looks like in the abstract. “Batch data comes in real time from sensors in the production process, and is reliably and cyber-securely transferred to an (edge- or cloud-based) analytics platform. Digital twin models process and use such sensor data to work out simulations and optimization scenarios and suggest ideal production parameters.
Then the communication goes in the opposite direction: if the model with the real-time data sees an optimization, the optimal control action flows back through the data platform to the production process and its assets.” Thus, in an advanced implementation, the digital twin can fully automate production itself, but as an intermediate step, the digital twin can also provide optimization suggestions to human operators.
“A twin does not have to be a 3D replica of the physical asset,” Canzani emphasizes. “The solution must replicate the process dynamics to optimize it. 3D models can be a comprehensive representation, contributing to usability, but at the core of digital twins are machine learning, data-driven and mechanistic models.”
Cognizant is betting heavily on true digital twins running on real-time streaming platforms. The initial focus is on large organizations in the life sciences domain. Think of pharmaceutical companies, who can use digital twins to optimize the complex production of drugs.
A model is never discarded
There are many challenges associated with the successful implementation of a digital twin. The creation of the model itself, is not the biggest. Stas: “Customers have often already built models, or at least a basic design. These models must then be modified or re-implemented to ensure real-time control. Communication must be automatic and fast enough, and ideally in both directions.”
read also
What is a true digital twin? Not a static but dynamic process
Canzani has another caveat: “Sometimes organizations have already developed an excellent model, but used a closed environment such as a third-party software to do so. A closed ecosystem creates problems, because then it becomes difficult to provide real-time connectivity with other platforms and automation systems.”
Regardless, organizations that already have a model, but not a true digital twin, have a foundation to start from. “The development of such a model is never wasted time,” Stas emphasizes.
Connecting systems and people
Much of the complexity comes from the connection between different teams that must work together on an end-to-end data flow for the digital twin. A digital twin has an IT component with an automation team going after the data and an analytics team building the model. In turn, the data comes from the OT environment. All those business domains are populated by different experts with different priorities and expectations. A successful digital twin must connect not only the systems, but to some extent the people as well.
“People need to communicate first to come to a common understanding, and then you can connect the technology,” Canzani said. “A twin is an end-to-end system, which needs to connect many different things. Many organizations are strong in one of three key areas; manufacturing, IT or analytics. When you bring the three together, the real added value is created.”
Reusable and scalable
“Ideally, you don’t develop a digital twin as a one-off project,” Stas says. “The solution can be scalable and adaptable to different processes. You build the twin for one production line, and later when you introduce a new product you can reuse the framework and optimize production from the get-go with the twin.”
Ideally, you should not develop a digital twin as a one-off project
Christophe Stas, Associate Director Life Sciences Manufacturing Cognizant
Cognizant developed TwinOps with that in mind. That is a platform that consists of building blocks for data ingestion, validation, transformation and storage. Those building blocks can be reused after validation, providing a sustainable and scalable link between data along one side and digital twins along the other. “TwinOps is not only a framework with best practices, but also a Python-based package of customizable components and building blocks for standardized, compliant and repeated workflows,” Canzani clarified.
Broad knowledge
Furthermore, according to Stas, Cognizant has a good understanding of the challenges across all domains involved in implementing a digital twin. “We know where the challenges are and the bottle necks are,” he says. “We can make a good assessment of whether the investment is worth its money, and what constitutes a realistic business case.”
Stas continued, “You need vendor agnostic knowledge for a good assessment of relevant technologies, with a great understanding of rules and standards, and of course cybersecurity.”
Worth millions
An investment in a true digital twin, will pay for itself in the long run. Canzani: “A digital twin can improve the yield of a process by ten percent. In the pharmaceutical sector, in some cases that amounts to optimization of 400 million euros per year.”
Still, don’t start big right away. “The beginning of every project is an assessment of the maturity of the organization,” Canzani knows. “In doing so, we look for a use case that is already advanced enough, where there is both data available and a good understanding of the underlying processes to model with the digital twin.”
“Then we build a digital twin for an initial limited use case,” Canzani continued. “We start on a small scale but build the twin right away with an architecture that supports real-time usage and scalability. The workflow must be automated and repeatable according to TwinOps best practices.” Such a small rollout will immediately add value and show that the investment is worth the money. “Based on that first small but complete use case, we build trust with stakeholders to clearly realize the feasibility and potential. That way we can scale up the project and roll out other use cases at the same time.”
This is an editorial, created in close collaboration with Cognizant.