In this editorial, Andrew Sherlock, director of data-driven manufacturing at the National Manufacturing Institute of Scotland (NMIS), argues that while AI hype often focuses on chatbots, the real transformation in manufacturing will come from sector-specific data-driven applications that enhance process control, design and decision-making, as long as manufacturers invest in robust datasets, workforce skills and practical integration on the shop floor.
Artificial intelligence (AI) is rapidly becoming a defining topic across the manufacturing sector. While much of the public discussion focuses on large-scale language models (LLMs) such as OpenAI’s ChatGPT, the impact of AI in the industry goes far beyond its role as a digital assistant or content generator.
A new wave of sector-specific applications is beginning to emerge, shaping the way manufacturers design processes, make decisions, and derive value from the data running throughout their operations.
This change comes as production systems become more complex, supply chains become more demanding, and expectations for efficiency and quality are higher than ever. Traditional tools are struggling to keep up, and manufacturers are being asked to consider how AI can support activities that have long relied on intuition and manual oversight.
Across sectors, projects are demonstrating how targeted AI models can improve accuracy, enhance process control, and provide clearer visibility on the factory floor. As momentum grows, the question is no longer whether AI will impact manufacturing, but how deeply it will penetrate engineering practices and what foundations will be needed to realize its full potential.
Data access is central to that progress. At the National Manufacturing Institute Scotland, researchers are developing new AI capabilities while also generating the datasets they need to be able to run them in real-world environments. Even the most sophisticated software relies on robust training data, and without it, its value quickly diminishes.
Practical examples are already emerging. NMIS works with major aerospace companies to model how parts behave during forging. This is a very complex process where some of the underlying physics are still not widely understood. By combining machine learning and targeted trials, the team is creating datasets that help predict changes in material properties, giving engineers the insights they need to more accurately design and optimize processes, reducing the number of trials, reducing waste, and increasing overall efficiency.
Innovation is also accelerating across the broader ecosystem, with more companies developing AI tools for design, simulation, and factory-level decision-making. For example, a UK software developer is helping engineers extract value from computer-aided design (CAD) data through the HOOPS AI platform. The HOOPS AI Platform is a framework that unifies access, preparation, and training of machine learning models using 3D geometry.
Emerging engineering platforms are also using AI-powered models trained on existing simulation data to reduce traditional simulation times from days to seconds. Engineers can refine designs in real time, speeding development and reducing reliance on lengthy physical and computer-based tests.
NMIS is also working with academic partners, including colleagues at the University of Strathclyde and researchers at the University of Edinburgh, to apply AI to factory operations. By capturing raw data streams from equipment such as forklifts, new algorithms can track parts, identify inefficiencies and bottlenecks, and highlight opportunities to improve flow and minimize downtime within production areas.
Despite progress, AI adoption on the factory floor is still in its infancy. Through the Data Driven Design and Manufacturing Collaboration project, part of the Glasgow City Region Innovation Accelerator Program and funded through Innovate UK on behalf of UK Research and Innovation, NMIS works with organizations to bridge the gap between manufacturing and digital technologies, giving engineers the skills and confidence to apply data-driven methods to their own businesses. To date, more than 120 projects across aerospace, energy, food and beverage, construction, and electronics have already demonstrated how a data-driven approach can reduce emissions, improve part accuracy, and increase reliability.
AI is already impacting the way manufacturers design processes and make decisions. As targeted tools continue to mature, the opportunity is to embed them more deeply into engineering operations and empower employees to use them effectively. This shift from broad discussion to practical integration is where AI will have the most tangible impact, helping manufacturers derive greater resilience and value from the data that underpins modern production.
