High-Quality Data is the Foundation of a Digital Twin

Building the next-generation energy grid will require utilities to embrace digital twin and Building Information Modeling (BIM) technologies. BIM is an interactive 3D design model of a building or other built infrastructure assets, such as a bridge or nuclear reactor. BIM models typically focus on the design phase of a project to enable stakeholders to analyze constructability and track progress.

Interactive 3D models for the operational phase of an asset or system of assets are commonly referred to as digital twins. Digital twins combine data from multiple sources, including the original engineering design, construction, inspections, and real-time operational data, including sensors. This article focuses on creating digital twins.

Utilities can use digital twin technologies to streamline design, construction, and O&M and to optimize performance with analytics and predictive maintenance. Advanced technologies, such as artificial intelligence and augmented reality can be built on these 3D interactive models.  

What is keeping the utility industry from implementing digital twins?

One of the primary barriers to implementing digital twin technology is the lack of high-quality data. A digital twin requires data from various lifecycle phases, including engineering design, product manufacturing, construction as-builts, and commissioning and test reports. All of this data exists, so why is it so hard?

Stakeholders throughout the supply chain create, store, and exchange data based on their unique needs and the needs of their downstream users:

• Engineers create CAD design drawings with Compatible Units (CUs) and a Bill of Materials (BOM) in a spreadsheet.
• Supply Chain purchases products with catalog IDs using unstructured item descriptions. Manufacturers provide PDF specification sheets using internal product IDs and their own terminology.
• Construction contractors create redlined as-builts on paper drawings using their own terminology and symbols.
• ERP systems store accounting data based on CUs and unitization models.
• GIS systems store data for geospatial analysis but don’t have complete engineering datasets.
• Engineering systems store full specifications but don’t have traceability data such as serial numbers.
• Other data, such as test reports and warranties, might be stored in a PDF format in a shared drive—and these PDFs might be scanned hand drawings!  

A digital twin needs all of these datasets to support modeling, analysis, and preventative maintenance. Extracting data from PDFs, spreadsheets, and hand-written forms is time-consuming and prone to human data entry errors.

While there are many gaps in the data flow from one phase of the asset’s life to the next, construction and as-builting are typically where the data flow commonly breaks down. There are two reasons for this:

• What gets installed in the field may be different than what was initially designed to account for constructability liability issues. All as-built changes require field users to precisely document as-builts with information about installed assets, including substitute materials.
• Field crews manually create redline as-builts on paper drawings under harsh environmental conditions. Data legibility, accuracy, and completeness can be compromised.

How can we solve the digital twin and construction as-builting gap?

Digital as-builting is the process of capturing as-built data in a way that allows data to flow from one phase of the asset’s life to the next—a Digital Thread. With a Digital Thread, as-built data can flow seamlessly from the field during construction into systems of record without manual data entry to ensure accuracy and fidelity. The key to digital as-builting is automated data capture that is feasible for use by construction field crews.

The ideal Digital Thread workflow would look something like this:

• Engineers create a digital graphic work design (GWD) with a digital BOM.
• Manufacturers label products with a traceability barcode that links to digital product specification data and other essential datasets.
• Construction crews create a digital as-built by updating the digital GWD with asset data, using automated capture with barcodes and GPS.
• Digital as-built data is delivered to systems of record such as ERP, GIS, OMS, and ADMS for streamlined integration.
• O&M crews leverage the traceability barcode to attach inspection, performance, and repair data to the correct asset in the system of record.
• Advanced digital twin and analytical tools leverage the systems of record populated with high quality asset.

This workflow promotes a continuous flow of data through an integrated Digital Thread from design to systems of record.

Steps required to implement a digital thread

The ideal workflow described above requires coordination and integration of multiple steps throughout the supply chain and the asset’s life cycle, including:

• Engineering GWD and the BOM need to be structured to enable linking to the manufacturer’s product IDs and digital specification datasets.
• Manufacturers must provide product specifications in a standardized format accessible from a barcode with a unique product ID.
• Construction crews need digital designs that can be updated easily with data from product barcodes.
• Digital as-builts need to be structured to deliver data seamlessly to various systems of record (and this might mean sending different datasets on the same asset to different systems).

Implementing the coordination and integration described above is challenging because it requires a holistic view of an asset’s lifecycle from design through decommissioning. It requires looking beyond the typical silos of engineering, supply chain, construction, and operations. It requires a start-with-the-end-in-mind mentality.

• Envision a digital twin that performs advanced analytics and provides meaningful actions using artificial intelligence and machine learning.
• Ask, “What type of data would this digital twin require?”
• Work backward to ensure the data is collected, stored, and exchanged in the design and construction phases.

Exchanging data from one phase of the asset’s life to the next requires coordination and sometimes standardization among stakeholders with different goals and incentives. A digital twin needs a Digital Thread that enables data to flow from design to manufacturing, construction, and operations.  

To support this vision of the future, we are leading a new IEEE work group focused on creating an industry practice for Supply Chain and Asset Traceability for Electric (SCATE). The end goal is to develop standards and technology that will enable manufacturers of electric grid equipment and materials to apply barcodes with unique IDs to their products to allow stakeholders along the supply chain to access product specification data in a standardized format. Asset traceability barcodes applied by the manufacturer are an important component in bridging the gap between design, construction, and operations.

Conclusion

The promise of digital twin and related technologies of artificial intelligence, advanced analytics, and augmented reality for energy infrastructure will only be possible with high-quality data. The industry needs to invest in the seamless transfer of data from design to manufacturing to construction all the way through decommissioning. Digital Threads and smart barcodes based on open industry standards are two key enabling technologies that can make this vision a reality.