Ask anyone in test and measurement what’s changed in the last few years, and AI is top of the list. But according to Allan Solomon, VP of NI EMEA Sales at Emerson, and Sean Fuller, Senior Regional Sales Manager and NI UK Site Leader at Emerson, the more interesting story is what hasn’t changed – and how that’s exactly what’s pushing engineers toward AI in the first place.
The real constant: pressure
Solomon has been watching this build for two decades. He remembers when a four-year automotive design cycle was standard. “Then it became one year,” he says. Now even aerospace, an industry historically insulated from tight deadlines, is asking how it can move at automotive speed.
“There used to be companies I’d go to where they’re like, ‘Oh, we don’t have time pressure, we just got to get this right’ and I swear that was especially true in aerospace,” Solomon says. “[But] it’s everywhere now. This pressure is consistent.”
Fuller sees the same dynamic but frames it as a shift in what companies are actually competing on. “Before, it was always about devices becoming more complex and engineering teams not having more time,” he says. “But now I’m seeing it shift – they’re talking about engineering productivity as the competitive advantage of the company.” That means getting engineers off repetitive, administrative work and back onto the problems only an engineer can solve.
Fuller points to Formula One as the clearest example. Teams used to spend months manually sifting through millions of data points from a single race. AI has collapsed that timeline, letting engineers get straight to “what really made that wing on the Formula One car better or worse.” With new regulations reshuffling the field this season, that speed has become critical: teams are using AI, including NI’s own AI assistant Nigel, to find the next tenth of a second before the next race.
That, Fuller argues, marks the real inflection point of the last 12 months: less “can I use AI?” and more “I absolutely need this specific type of AI to solve this specific problem.”
Where the investment is actually landing
Both Solomon and Fuller point to semiconductors as the sector moving fastest – and for a practical reason. “They’re trading pennies for chips, if you will,” Solomon says, “so they’re always a little bit on the leading edge of new technology integration.” Semiconductor companies were building data-enabled workflows before AI arrived and have been quickest to fold AI into them since.
Defence, by contrast, is “the most wait and see,” Solomon says, largely down to security concerns, though he’s careful to note this varies by organisation rather than applying evenly across the sector.
Fuller sees the UK’s semiconductor position as a genuine structural advantage. “We’ve got world-leading universities, and we’ve got world-leading startups leading into industry leaders such as Pragmatic Semiconductor,” he says. He points to chip design itself as a case in point – AI being used to help route the internal wiring of chips containing billions of connections, optimising for space, performance, and speed simultaneously.
Beyond the AI headline: model-based engineering and digital twins
One of the more grounded points both executive make is a caution against over-indexing on AI alone. “I feel like right now we’re so focused on AI, and in reality, I think all the pressure has caused other technologies or engineering approaches to come to the forefront,” Solomon says. He cites model-based engineering – adopted by companies under pressure to get new designs validated faster – and a rising uptake of digital twins and simulation, neither of which is inherently AI-driven but both of which are being accelerated by the same forces pushing AI adoption.
What it actually means for engineers day to day
Strip away the strategic framing, and both men describe a similar shift in daily engineering life: less logging, more thinking.
“It’s a more fun and rewarding job as an engineer,” Fuller says. “You’re taking away the bits they don’t really want to be doing – the repetitive tasks, logging something in Excel, typing up data. That’s not what they became an engineer for.” The result, he argues, is faster prototyping cycles and more time spent on what drew people into engineering in the first place: building a better product.
Solomon agrees, but adds a caveat that’s easy to miss in the optimism: AI adoption creates new work, too. “As an engineer uses AI, they have a new task called quality control,” he says. He traces this back to a single root cause – an “exponential increase in complexity” in the products themselves. A metering company that once made a largely mechanical product now ships meters filled with RF chips. Cost-reduction targets haven’t gone away; they’ve just been layered onto far more complex hardware.
What differentiates AI from other engineering tools, in Solomon’s view, is the shape of the payoff. “AI gives us an exponential solution versus a more linear solution,” he says. “AI just offers this future that could be so powerful in combating the complexity. I think everyone’s enamoured with it.”
Will AI keep pace with rising product complexity or start driving it?
Asked whether AI can keep up with ever-increasing product complexity, Solomon reframes the question entirely: he doesn’t think AI will simply keep pace – he thinks it will start setting the pace. Pointing to the current wave of advanced robotics companies as an early signal, he argues AI-assisted decision-making – even at the relatively early ‘advisor’ stage represented by Nigel – is already surfacing solutions engineers wouldn’t have reached alone, and that this capability will increasingly drive product design rather than just support it. Fuller agrees without qualification: “It drives the innovation in the end. The speed of AI and the evolution of that is incredible.”
Where Nigel – and the wider market – goes next
Fuller maps a clear trajectory for AI in test and measurement, using Nigel as the reference point: from an advisor answering basic questions, to an author generating code, to (eventually) an agent capable of taking a task, breaking it down, and running an entire workflow with minimal prompting. “I think the impacts, the things that it will solve, is going to be exponential,” he says, though he expects the real gains to depend on how quickly organisations work through security and compliance requirements before “unleashing” Agentic AI at scale.
Solomon’s vision goes further still: a future where an advisor can take a test specification straight from R&D, build out requirements, and recommend the hardware, software architecture, and data analysis needed – walking an engineer through the entire design process. “I believe that’s going to become real, and it’s going to become real and be faster than any of us think,” he says.
For Solomon, the deeper shift isn’t really about the tools at all. “I think for engineering, it’s going to become a little bit less about the tools and the how, and a lot more about the output and your ability to have a vision of how you want to solve something.”
Humans stay in the loop – for now, and maybe always
Pressed on whether AI could eventually operate without human oversight, both were clear that autonomy has limits. “I always see a future where the engineer is always in control and overseeing things,” Fuller says. Engineers may hand AI more autonomy over specific tasks, but a human still needs to “verify and validate that what’s actually been built and what’s been decided is creating the right thing.”
Solomon reaches for a historical parallel: the disappearance of rooms full of human accountants once spreadsheets arrived – followed by an entirely new set of roles nobody had predicted. “I think the role of humans will continue to evolve, and I doubt – I hope – we’re not living in a world that’s run by robots and AI.”
Not quite Terminator, then. But a test and measurement industry where the engineer’s job looks less like data wrangling and more like judgment – backed by tools that are only just getting started.
