Once you have a digital twin of a plant that includes predictive and prescriptive AI capabilities, how could you improve it? TwinThread is doing that by adding genAI and agentic AI in Advisor to help accelerate model interpretation, as well as a new solution aptly named Perfect Centerline to help manage and stabilize the process effectively. These can both improve time-to-value for customers.
Additional AI: Advisor
The TwinThread Industrial AI platform is founded on pulling in operational data to build a comprehensive digital twin. To optimize performance, this PaaS includes predictive AI (as machine learning), and to take it a step further and close the loop, it also includes prescriptive AI. In addition to the core platform, TwinThread offers
- a virtual operations center enabling a manufacturing engineer anywhere to support the operation
- pre-built solutions for an even faster start.
You can find the basics about TwinThread and its offerings in our Insight posted in May 2025.
What’s new in this release is TwinThread Advisor, which includes both GenAI and Agentic AI capability. These language-based AI capabilities are context-aware, helping people interact with and interpret the results from the numbers-based predictive and prescriptive models in the core platform. This, in turn, can accelerate time to value, as anyone who has used AI to leverage a vast pile of information has experienced.
Incorporating the Advisor on top of TwinThread’s rich data set in a complete digital twin live model of a production site offers many potential benefits. Supporting understanding is a foundation. Process engineers, operators, and digital teams, as well as data scientists, can all benefit. It enables Q&A in their native language, performs global search, and builds analytical content. The agents in Advisor can integrate with various systems, and each has a specific persona that works in conjunction with other agents in a context-aware manner.
Advisor can use the TwinThread knowledge base of that customer’s operation, plus, if desired, the Internet. The customer has complete control of where the system pulls data. Using the corporation’s own data only is an option that can actually reduce spending on the cloud hosting service. It has complete security, is designed for privacy, and does not entail the sharing of one customer’s data with others. Advisor is currently available as a preview, an optional capability.
Perfect Centerline
Engineers in process industries are almost certainly familiar with centerlining to minimize variability from run to run. Most companies do it, but this new pre-built TwinThread Perfect Centerline solution automates managing the centerlining process and improves it with each run. These advanced capabilities are now fully launched to make it simpler to derive and manage centerlines with confidence.
The customer can set up to five target KPIs. The system will then monitor and automatically optimize to those objectives. The system enables overriding automated setpoints and displays the centerline, current readings, and statistics about the run in real-time. To avoid users being overwhelmed, the system can also simply display centerline deviations. TwinThread projects a 50-75% increase in process stability, which can yield significant benefits.
TwinThread expects this to be a point of entry for most customers, as a baseline. The process model first derives a centerline, which process engineers can modify, if desired. Once the data is all connected in the platform, centerline analysis can be performed in minutes, indicating how much the process might improve. If the financial impact per percentage of each objective KPI is established upfront, the system can deliver a value-based view of optimizing to the centerline. TwinThread states that if the company has a historian, this transition from legacy to advanced approach may even occur within a week.
Deeper Insights for Value
Once implemented, TwinThread’s proven predictive and prescriptive industrial AI and digital twin platform has a comprehensive and up-to-date data set. This is a rich starting point for genAI and agentic AI to deliver value. It is also a perfect home for advanced centerlining capabilities.
Thank you, Elise Loffredo, Jason Dietrich, and Andrew Waycott, for briefing Rick Franzosa and Julie Fraser.
on these new capabilities. We are excited to see both the added AI and the streamlined centerlining. We look forward to learning what you add next.
