Jump to section
In 2018, someone at Smith+Nephew put a Tulip factory kit on a credit card. It cost a couple of thousand dollars, and nobody above the plant floor approved it. This year, Smith+Nephew's Suzhou factory won the Shingo Award, the operational excellence world's gold medal.
At Operations Calling 2026, Sarath Krishnaswamy told the story of what happened in between. His story is about technology, but mostly about the people who made it work: operators, maintenance techs, quality leaders, and a small OT team that was willing to be wrong in public.
A company that can't afford mistakes
Smith+Nephew is a $6.2 billion portfolio medical technology company with about 17,000 employees. Its products reached 15 million patients in 2025. It started 170 years ago making bandages in the UK, and it now makes hip and knee implants, robotic surgery tools, sports medicine products and advanced wound dressings.
That range makes manufacturing complicated. One network runs high-speed bandage lines, pharmaceutical and biologics production, orthopedic machining that has barely changed in decades, and sports medicine work that still involves tying knots by hand. All of it is regulated by the FDA and other agencies. As Sarath put it, there is no room to let a mistake leave the floor.
The company's purpose is Life Unlimited: removing the limits that keep people from living fully, like dancing at your child's wedding. Its cultural pillars are care, courage and collaboration. Sarath's team added four tenets of its own: Be Valiant, Build It, Own It, and Elevate People.
The first lesson: the dashboards that didn't land
The OT team began with a simple idea. Collect machine data, show people how the plants were really performing, and the insights would sell themselves. By 2021 they were scaling hard, with Tulip stations on lines across the network and dashboards that showed in bright red how far behind everyone was.
The floor pushed back. IT hurdles slowed rollouts. And when the team put a belting process on a knee-finishing line, operators were flagged for skipping steps. The team's first reaction was to blame the operators. They learned the operators were being wise. Experienced hands knew a part was already good enough to skip the rough finishing step and save time, and the OT team had never watched closely enough to notice.
"We tried scaling the way we knew," the team concluded, "but we couldn't catch fire on the shop floor." These were seen as "OT projects" when OT was about five people. They needed to be site projects.
The turn: let the people build
The breakthrough came from Suzhou, where the team asked a different question: what if the people closest to the problems built the solutions? They launched User as Developer, a program that taught anyone to build a micro app. Operators and supervisors joined, and so did people from finance and HR.
The first cohort had 17 people, and all 17 finished. Together they built 15 apps. Several are still in use. One was a maintenance tracker built by a maintenance technician. It has been revised about 150 times, and it is now the most used app in the entire network, six years later. It will likely become the foundation of the company's next-generation maintenance approach.
Then came the moment Sarath says he knew it was working. At a review session in late 2022, he saw the company's top 10 apps and top 10 users. He didn't recognize the first four apps, which had Chinese-character titles. His own team's first appearance was at user number three. By the end of 2022, six of the top 10 apps were not built by OT.
Sarath had a sharp response to the "Wild West" worry that comes up at every conference. "This is what's happened," he said, when you let people build. The OT team moved to the background to build guardrails, app standards and a common architecture, so every site's choices would fit together.
Small questions, big returns
Each site chose its own direction: Suzhou on connectivity, Memphis on dashboarding, Costa Rica on package-to-export, Mansfield on incoming inspection. Because everything sat on one architecture and one data model, every answer became the foundation for the next. And once people had the information they needed, they kept going.
Suzhou, the slitting plan. The team asked for a simple way to allocate material. Five hours of planning became about one second, and roll utilization rose from 94% to 99%. The business case promised $90K in savings. The reality was closer to $500K, and the app is now being copied to factories in Hull and Milton.
Costa Rica, package to export. The ask was to track order paperwork. Shipping inventory dropped 70%, consolidation went from two weeks to one, and sterilizer utilization became measurable for the first time. That led to a new idea: since the EtO chamber costs the same whether or not it's full, work in process could be used to fill it.
Costa Rica, scrap at the line. The site leader was skeptical, so Sarath offered to pay for the pilot and pull the stations out if they didn't like it. An operator saw scrap climbing on one assembly, called over two supervisors, and together they traced and fixed the source. About $100K in scrap was cleared, and the app paid for itself in three days.
Nobody had predicted this value in a business case, not the OT team and not the associates on the floor. Their people uncovered it by doing the work.
Quality as the champion
The OT team had stayed away from the GMP space, worried about how Quality would react. They were wrong. Quality had wanted an electronic device history record (eDHR) for years, and they were excited to work together.
Operations, focused on short-term ROI, found it hard to fund connectivity across the floor. A Quality VP became the champion who got the Memphis pilot approved, on the promise that it would make FDA audits easier and that savings would follow. The only benefit the plant would initially sign up for was reduced paper and off-site storage. Sarath's team called the project their Trojan horse: they wanted eDHR, but they really also got the connectivity underneath it.
The project also gave birth to a new governance model, the Manufacturing Tech Council, where OT, IT and Quality govern deployments together and advocate to executives as one voice.
The Memphis pilot delivered:
$398K in first-year benefit, against $227K promised
Zero documentation non-conformances, on a value stream that previously cost about $400K a year in them
97.7% first-pass yield, up from 94.4%
The Memphis manufacturing leader, who had worried that the extra scans would slow the line by a few seconds, said he had to "take some of that back" because the trade-off was "HUGE" in the analytics.
Lighthouse sites and a three-month test
In 2026 the team formed a lighthouse working group with the leaders and OT leads of its three most forward-thinking factories: Suzhou, Costa Rica and Penang. The biggest lesson from Cindy, Jose and Fred, the site leaders, was the same. They established people and culture first, because their people were curious, wanted to build, and were allowed to.
To test the approach, the team built one app product, covering tiered accountability, 5S assessments and Gemba walks, with Tulip implementation engineers. They piloted it in Costa Rica on July 24, exported it to Memphis on August 14 and to Hull on September 4. It went from build to three sites in about three months, with a common core and site-shaped edges. It will reach a dozen-plus sites by the end of the year or early next.
Next up: AI, with the same approach
The latest step is the jump to AI tools. Weeks before the conference, Tulip ran a two-day AI workshop in Costa Rica using real site data. Sarath's team heard the same thing they'd heard about low-code years earlier. Mario Rios, OT Lead for the Sports Business Unit, said these tools are "simply an evolution of what Tulip is about: democratizing tech to deliver value to the shop floor." Another participant noted that an agent could spot issues from just a few tickets, instead of relying on slow back-and-forth conversations with users who are "suffering" in the moment.
The vision from 2020, a connected digital factory network, is now in sight. The path runs from digital silos to connected plants to a predictive network, with AI agents helping people magnify their domain knowledge.
Two lessons to take home
Being valiant still matters; it's just who you fight for. "We were fighting for data, now we're fighting for people." The team champions the front line, talks to users constantly, and tries to live their lives.
You can't have AI unless you have I. Without real intelligence about what your operation does, and the wisdom that only people have, AI tools won't do anything. As Sarath put it: "If you don't have that, don't bother us with the tools."
His team's story proves the reverse is also true. Start with curious people, give them guardrails and room to build, and the tools, including AI, become a power drill in the hands of people who already know where to drill.
Want more stories like this one?
Smith+Nephew’s story is one of many from Operations Calling 2026, where manufacturers shared how they’re building adaptable, AI-ready operations with their people at the center.