AI is Coming for Test Engineers, But This GCC Says Humans Aren’t Going Anywhere

When generative AI entered software development, much of the conversation centred on whether AI will replace programmers. However, executives at National Instruments (NI), Emerson’s fully owned test and measurement business unit, believe the next disruption is headed towards test engineering.

At NI Days India 2026 in Bengaluru, Emerson unveiled the latest capabilities of Nigel AI, its AI assistant for test engineers, adding code-generation features while outlining plans for agentic workflows expected in early 2027. But despite the rapid advances, the company insists that engineers will remain central to every critical decision.

“AI is not a trend or a commercial off-the-shelf technology,” Jared Aho, Senior Director of NI Marketing at Emerson, told AIM during the event. “We’re going to figure out how to abstract those complexities, bring it into our platform, and make it available to engineers, scientists, and students to become more productive.”

The distinction is that test engineering extends far beyond writing code.

“People immediately think AI means coding. But test engineering is greater than coding,” he said, referring to various other functions like requirements, system design, validation, security, data analysis, and much more.

That philosophy underpins Nigel AI’s roadmap. The tool began as an ‘advisor’, helping engineers understand documentation and recommend testing approaches. It now adds an ‘author’ mode capable of generating LabVIEW and TestStand code to build individual test blocks and control test flow, while the next phase will see it orchestrate larger engineering workflows—from reading product requirements to generating complete test plans.

Emerson completed its $8.2 billion acquisition of NI in October 2023, bringing the test and measurement company into its Software and Control business and expanding its presence in high-growth sectors such as semiconductors, electronics, transportation, aerospace, and defence. 

Humans Stay in the Loop

Despite the industry’s growing fascination with autonomous AI agents, Emerson says it has no intention of removing engineers from mission-critical workflows.

“The engineer is still ultimately the one responsible,” Aho said. “We see no future where the engineer is out of the loop.”

The approach becomes particularly important in industries such as aerospace, automotive and semiconductors, where software errors can have real-world consequences. 

Baskar Ceri, Managing Director of Emerson Test & Measurement India, echoed the point to AIM. “Our AI is always engineer-in-the-loop or human-in-the-loop. Some industries will adopt faster than others, but in aerospace and defence, adoption will naturally be slower because the human remains responsible.”

Instead of replacing engineers, the company believes AI will elevate them.

One internal example cited by Aho showed a task that previously required around 20 hours over two weeks being completed in 20 minutes using LabVIEW together with Nigel AI. The company also said customer studies over the past year indicate users have reduced test development time by around 50%, with the figure based on market research and user feedback rather than internal benchmarks alone.

Measuring AI’s ROI is the Wrong Question

As enterprises race to deploy AI, executives increasingly demand measurable returns on investment. Emerson argues that organisations may be asking the question too early.

“Companies are mandating AI,” Aho said. “Then immediately they ask us, ‘What’s the ROI?’ But we don’t even yet know the right places throughout test engineering where AI delivers the most value.”

Instead of focusing solely on code generation, he believes AI’s biggest gains will come from reducing engineering bottlenecks across the entire testing lifecycle. Ceri pointed to examples where Emerson’s existing test platforms had already delivered measurable gains, including fourfold improvements in testing time and tenfold reductions in infrastructure investment before AI was even introduced.

With AI, he expects similar improvements to come from productivity rather than outright labour replacement. “A task that earlier took ten man-days might now become half a day,” he said. 

He also expects AI to transform validation itself. Rather than engineers manually writing thousands of test scenarios, AI could automatically generate millions of permutations, pushing validation coverage significantly higher.

Building AI, Not Just Consuming it

The conversation also turned to India’s role inside Emerson’s global engineering organisation.

While many multinational engineering centres in India have historically executed work developed elsewhere, Ceri argued that the company’s Bengaluru global capability centre (GCC) has evolved from a capability centre into an innovation hub with ownership of core products. 

NI entered India in 1998 with a liaison office in Bengaluru before converting it into a wholly owned subsidiary in 2000 and establishing an R&D centre soon after. In 2022, the company further strengthened its India presence by opening an Engineering Innovation Centre in Bengaluru to support customers, startups, and engineers working in aerospace and defence, while providing facilities for prototyping and training.

Ceri cited TestStand as one example of software whose development is led entirely from India.

The company is also expanding its engineering presence while investing in academic collaborations, startup programmes, and hackathons to broaden access beyond India’s largest cities.

Emerson is partnering with IITs, IIITs, NITs, L&T Technology Services (LTTS), Tata Technologies, and AWS to reach more than 700 engineering colleges through an AI-on-the-edge hackathon focused on aerospace, transportation, and intelligent manufacturing. 

On the commercial side, Emerson is expanding its distribution network and system integrator ecosystem to increase adoption across tier-2 and tier-3 cities as well. The company recently added Aimil alongside existing distribution partners such as VVDN and element14 to widen its reach. Aho said the partners would help Emerson work more closely with SMEs beyond major technology hubs.

Nigel is Taught, Not Trained

One subtle distinction the company repeatedly emphasised was its choice of language around Nigel AI. Employees are discouraged from saying the assistant is “trained”. Instead, Emerson says Nigel is “taught”.

“It’s built on best-in-class models,” Aho said. “Training implies you’re building a model. We’re augmenting an existing model with our engineering knowledge.”

On why the company is not building its own model, he said Emerson simply doesn’t see value in competing with frontier model companies given the extensive capital required. According to a PitchBook report, AI startups globally raised $255 billion in Q1 2026 alone.

So instead of building a model, Emerson focuses on adding test-and-measurement expertise on top of existing ones. Nigel uses commercial LLMs, including OpenAI’s GPT models, alongside retrieval-augmented generation (RAG) to incorporate NI’s technical documentation, knowledge bases and product literature.

“As we go forward, you’ll be able to pick whatever model. The future on that one will be as we get further down into next year; there will be a selector: what model do you want to use?” Aho said.

The system deliberately excludes community discussion forums because Emerson cannot verify the accuracy of user-generated content. It also avoids competitive comparisons between NI products and rival hardware, reflecting guardrails intentionally built into the assistant.

Aho explained that heavily regulated industries such as aerospace and defence are asking for deployments outside the public cloud. Initially, Emerson will rely on cloud providers’ secure government/private offerings before allowing customers to host Nigel in their own environments.

He also added that Emerson encrypts customer prompts before they reach Nigel and that no human can read them, but once the data reaches the cloud, the company relies on Microsoft’s security guarantees. “We are reliant on the security guarantees that Microsoft gives us.”