A: Just a year ago, we released what we call the DeltaV Edge, which is a new product in the DeltaV portfolio. It's a node that sits on the DeltaV system and allows DeltaV data to pass through a one-way communications protocol—or even a physical data diode, if you like—to populate another set of processors running edge-based technology. It’s a node that uses what's called HCI, or hyperconverged infrastructure, which is a next-generation version of virtualization. What this node lets us do is create an entire digital twin of the DeltaV system, not just the process data, but alarm data, event data and configurations—essentially the complete context of the DeltaV data on the open edge side of that node.
Once you’re there you have access to all the more modern edge analysis tools. We added new capabilities for exporting the data via standards like JSON and MQTT, in addition to the OPC standards that we often use in the automation world. Another part of the Boundless Automation vision is allowing control to actually run on edge platforms, instead of only on purpose-built hardware. And we’ll have our first releases of that by the end of the next calendar year.
Q: It’s certainly no secret that the industries Emerson serves are facing a bit of an expertise crisis, with the waves of experienced operators and engineers retiring. What role does artificial intelligence, especially in the form of generative AI, promise to play in complementing the more traditional, analytics-based advisory tools like Plantweb Insight?
A: First, I’d remind your readers that industrial automation was AI long before AI was cool. Within AspenTech especially, the vast majority of their portfolio is around optimization, advanced control, planning and scheduling. That's all built on data models, but also first-principles models used in the design and simulation of process plants. What’s different from the AI we were using 25 years ago is that instead of using a narrowly defined algorithm trained on a specifically relevant set of data, generative AI uses a general-purpose model trained on broad data sets to answer a whole variety of questions.
In terms of the products Emerson develops and the value we can deliver to our customers, I think in terms of three buckets. The first is in the area of customer support: anyone who supplies an industrial or consumer product is working on developing an interactive copilot to guide its users. The second area is in the configuration of products. Industrial products, especially software and systems, require considerable site or application-specific configuration. We actually have a product today that can take configurations of older, competitive control systems and using AI quickly translate them into modern configurations for DeltaV. But I think the AI application that stirs everyone's imagination is the idea of an actual operations copilot that you can talk to—an interactive operator or operations advisor that you can ask for input on anything.
Q: I’m looking forward to connecting in person again next spring when the Exchange community will be gathering in San Antonio from May 19 to 22. Other than moving from what has usually been a fall cadence to the spring, any particular innovations and new things we should look out for?
A: Exchange attendees get to see in action all the things we just talked about—and more. But most of our users who participate say the biggest takeaway is the interaction with the other users they meet there, many of whom are facing the same challenges they are.
