The current technology focus is all over the place and leads to many questions. My response to technology solutions that don’t address real business issues has been the same for more than 20 years. I ask, “That’s the answer, what’s the question?” You might remember me asking this same question in one of my previous blogs, Making Secure Technology Choices in the Energy Industry. In fact, this was a big topic on the panel I recently participated in as part of the program for the ARC Industry Forum 2021, titled ML and Analytics in Operations and Maintenance. But I digress.

Let’s take digital twins as an example. There can be many digital twins - there is not one exclusive digital twin. It depends on what you are trying to accomplish using that digital twin and knowing that may answer the questions you pose. 

A digital twin is a virtual model of a process, product or service represented in computer code, logic, equations and algorithms. A digital twin allows conditions and changes to be simulated via a model that recreates a virtual, real-life behavioral experience without exercising the actual equipment. In reality, digital twins are aligned with different functions that contribute to the overall equipment effectiveness (OEE) in manufacturing plant operations. There is not one, but many digital twins each for a different purpose or function. 

In the digital world, a particular digital twin can run hundreds or thousands of trials for a more thorough exploration of capabilities and options without using the equipment itself. Also, such trials operate remotely, in a completely safe environment without endangering environmental conditions. At a time in which there is such a strong focus on sustainability, this is no small consideration. In fact, AspenTech has been in the business of creating digital twins well before they were ascribed the “digital twin” moniker. They were the computer models set up for the design of and simulation services in chemical processing equipment. 

 

In addition to design, digital twins may be created to:

 

• Explore the possibilities of different operating plans and schedules to determine the optimal product manifest. 
• Model the errant behavior of processes and mechanical equipment allowing probability predictions of future performance and permitting the application of prescriptive advice to avoid or mitigate an impending situation. 
• Simulate the full system performance of a processing unit or a full plant and provide details of changing operational conditions, external influences such as weather, supply and distribution to be able determine the preferred cost and risk of alternative decisions. 
• Simulate and exercise the process for optimal production under the current operating conditions. 
• Exercise hydraulic models to predict and avoid distillation column flooding.
• Anticipate and prescribe action to avoid quality and yield issues.
• Validate instrumentation and recommend recalibration with inferential models.

 

Again, there’s no single digital twin that can address all the foregoing scenarios. Determining the right one to use requires plant personnel to truly think about the problem they are trying to solve, and get to heart of the matter by again asking, “That’s the answer, but what’s the question?” 

 

To learn more, read our white paper, The Digital Twin and the Smart Enterprise.