The National Manufacturing Institute Scotland (NMIS) has successfully collaborated with engineers on the factory floor at Diodes Incorporated in Greenock to unlock faster decision-making, reduce manual analysis, and improve efficiency in semiconductor manufacturing. Diodes is based in Texas in the United States and listed on NASDAQ.
The initiative, delivered through NMIS’s Data-Driven Design and Manufacturing Colab (D3M-Colab) programme, is helping manufacturers build in-house data science capability and apply it directly to operations.
At Diodes’ 24-hour wafer fabrication site, engineers previously relied on manual processes, using spreadsheets to extract and analyse information. While effective, this approach limited scalability and the speed at which findings could be made.
Semiconductor wafers underpin technologies across energy, transport and advanced manufacturing, with the UK market valued at more than £10bn in 2024 and projected to grow significantly over the next decade.
Gary Yu, President and CEO of Diodes, stated recently that 2025 ended with revenue growing by 15 per cent year on year, reaching the highest level of annual growth since 2021. Fourth quarter revenue was $391.6m. This has been with growth in the automotive, industrial and computing markets.
He said this “further highlights the success of Diodes’ design win initiatives and content expansion over the past year. We have continued to see demand improvements across all target markets and geographies.”
Working with NMIS, which is operated by the University of Strathclyde, Diodes’ engineers have adopted more advanced data science techniques, including Python and specialist semiconductor software. This shift is allowing greater automation of data workflows, reducing manual input while improving the consistency and accessibility of analysis.
The D3M-Colab programme included a 12-week structured training element, designed to equip participants with the skills to develop and deliver data-led projects independently, alongside ongoing support from NMIS data scientists. It has also supported the creation of dashboards that present data more clearly, enabling faster and more informed decision-making.
Andrew Burns, engineering manager at Diodes Incorporated, said: “Working with NMIS has given our engineers the tools and capability to move beyond manual data analysis and start using more advanced approaches. We’re now able to present our data in a much more accessible way, which means decisions can be made more quickly and with greater confidence. It’s helped us identify where we can improve efficiency, quality and cost, and it’s opened the door to how we can use data even more effectively in the future.”
Diodes’ engineers are exploring applications, including the use of artificial intelligence (AI) to analyse manufacturing outputs at scale and predictive maintenance to detect early signs of failure. These approaches could reduce downtime, lower maintenance costs and extend equipment life.
Andrew Sherlock, a director of NMIS, said: “The real value of this work is in helping manufacturers move beyond collecting data and start using it to address challenges in their operations. By bringing together engineering expertise and data science, we’re supporting teams to identify where improvements can be made and giving them the skills to turn that insight into action, driving better decisions and outcomes across the business.”
Scotland’s semiconductor capability is being enhanced and includes the soon to be opened NMIS National Advanced Semiconductor Packaging and Integration Centre (NASPIC) in Renfrewshire, which will support the development and commercialisation of next-generation technologies.
Led by Innovate UK on behalf of UK Research and Innovation, the Innovation Accelerator programme is investing £100m in 26 research and development projects to accelerate growth across three clusters in Glasgow City Region, Greater Manchester and West Midlands.