Cummins stakes out a people-led path to AI adoption
Darlington’s best-known engine maker, Cummins, is advancing a company-wide use of artificial intelligence with a clear message: the technology will assist rather than replace its workforce. Speaking about the approach being introduced across the manufacturer, Dom Coles — the firm’s services global continuous improvement leader — said the programme is intended to accelerate decisions and sharpen analysis while ensuring that employees remain firmly in control.
“We’re using AI technology to help the human beings in our organisation be quicker, faster, smarter – to get more decisions made more quickly.”
Coles, who has spent 13 years with the Darlington-based company, drew a comparison with collaborative robots already familiar on factory floors, stressing that AI will operate as a tool alongside staff, not above them.
“We’re seeing AI within Cummins on a very similar scale to the cobots you’ve seen on the shop floor… It’s human‑led every time.”
Checks, balances and a person in the loop
Addressing public anxiety around automation, Coles was explicit that the manufacturer will keep critical decision-making with people. He set out a framework in which algorithms can surface patterns and options, but proposals do not progress without human scrutiny.
“AI can recommend and accelerate workflows, but humans remain accountable for consequential decisions.”
He emphasised that the company’s philosophy rules out any automated pipeline from model output to product release.
“So it’s never going to be a case of we put it into the model, AI spits something out and we simply go ahead… There’s always going to be a person in the loop as we roll out a three‑strand approach.”
While Cummins has not publicly detailed that three-part rollout in full, Coles indicated that the firm’s engineering teams are already exploring where AI can speed design cycles and sift data more effectively to drive product improvements.
What AI will and won’t do at Cummins
- Use machine learning to analyse engineering and operational data more quickly, helping teams spot trends and options they might otherwise miss.
- Support faster, better-informed decisions by providing recommendations and accelerating workflows.
- Retain human accountability for safety, quality and market release decisions, with formal checkpoints before actions are taken.
Coles underlined that AI will act as an assistant to skilled staff. He sees its value in cutting through data-heavy tasks, freeing engineers and operators to focus on problem‑solving and judgement calls where experience matters most.
| Role | Humans | AI systems |
|---|---|---|
| Decision-making | Accountable for consequential choices | Provide options and evidence |
| Design & analysis | Define objectives, verify results | Accelerate modelling and detect patterns |
| Governance | Set checkpoints and standards | Operate within rules and guardrails |
Implications for Darlington’s advanced manufacturing base
For Darlington, where Cummins has a long-standing presence in engine production and engineering, the move underscores the town’s role in the UK’s transition toward digitally enabled manufacturing. A human‑centred rollout suggests continuity for skilled jobs, alongside demand for new capabilities in data analysis and AI oversight. It also signals a pragmatic route to productivity gains — applying algorithms to speed up number‑crunching and simulation while preserving the judgement that comes from experience on the shop floor and in design offices.
Coles’ remarks hint at the areas likely to feel the earliest effects: product development, quality assurance and continuous improvement functions that already rely on large quantities of operational and testing data. By tightening feedback loops and surfacing insights more quickly, the company expects to refine products and processes without short‑cutting the checks that underpin safety and reliability.
Reassurance on jobs and responsibility
In an era when automation can trigger understandable worries over employment, Coles sought to reassure that the company’s strategy is built around augmenting staff rather than supplanting them.
“It might be able to offer viewpoints and find patterns we may have missed, which could ultimately help us produce better products – but there’s never going to be a path straight to market.”
That stance places accountability and judgement with people at every stage, an approach that aligns with the rigorous governance expected in safety‑critical engineering. It also reflects the reality that AI models can produce errors or omit context, making expert oversight non‑negotiable.
What to watch next
As Cummins advances its three‑strand plan, the key markers for Darlington to watch will be the spread of AI‑assisted tools into daily engineering and factory routines, the emergence of training to support staff in using them confidently, and signs of measurable improvements in design and production cycles. For now, the message from the company is unambiguous: technology will be harnessed to make people “quicker, faster, smarter”, while people remain responsible for the outcomes.
