Every manufacturing leader today hears the siren call of artificial intelligence (AI) and automation. The potential rewards - predictive maintenance that prevents breakdowns, robots that boost productivity, analytics that streamline operations - are real and significant. However, amid the rush to implement the latest technologies, there's a critical caution often overlooked: without a solid foundation of basic conditions, high-tech solutions will likely falter.
The allure of AI vs. operational reality
Despite heavy investments, many AI and automation projects struggle to deliver meaningful returns. A recent survey found that only 30% of companies' digital transformation efforts produced a significant bottom-line impact - not due to poor technology, but because the underlying processes weren't redesigned and optimised. Gartner analysts estimate up to 80% of industrial AI projects never reach production, largely because companies try to "digitise" chaos, automating flawed processes or feeding AI poor-quality data.
The lesson? Technology applied to a fundamentally flawed process will simply amplify the flaws. As Bill Gates famously quipped, "Automation applied to an inefficient operation will magnify the inefficiency." If your current workflow is a tangled mess of workarounds and firefighting, layering AI on top will only help you make mistakes or generate waste faster and more expensively.
Fix processes and machines first
On the process side, getting the basics right means ensuring that work is done as it is written - standard procedures are documented, understood, and actually followed on the shop floor. On the equipment side, it means machines are kept in optimal condition - properly cleaned, calibrated, and maintained, with no 'quick fixes'. This kind of operational stability is the bedrock on which any advanced automation must stand.
It may not sound glamorous, but the impact can be huge. One World Economic Forum case study described how a UK process industry plant, after mapping and streamlining its value streams, achieved annual savings of £3.2 million, cut planned downtime by 24%, and reduced energy use by 24% - all before introducing a single new AI tool or automation system.
Crucially, many problems that managers think require advanced solutions might disappear once you restore the basics. An automotive industry study described a simple but striking rule: 8 out of 10 problems on the shop floor are resolved just by returning to standard operating conditions. Before investing in an AI-powered "root cause analysis" tool, first ask: have we tightened all loose bolts, calibrated the sensors, trained the staff on the proper procedure, and removed the clutter and scrap? If not, do that - in most cases, the issue will vanish without a single line of code.
Engaging the frontline
Who is going to restore those basic conditions and keep them in place? Not a remote data scientist, nor a corporate strategy exec - it's the frontline operators, technicians, and supervisors on your shop floor.
Frontline workers often have a deep intuitive understanding of where the "real" issues and workarounds lie. Rather than bypassing or ignoring this tribal knowledge, smart leaders tap into it. Including frontline employees in continuous improvement efforts yields more practical, workable solutions - and drives higher engagement.
Unfortunately, many companies have neglected this aspect. A 2023 BCG study found that only 14% of frontline workers had received any AI-related upskilling, versus 44% of senior leaders. If we introduce advanced systems without bringing our workforce along, we'll have fancy technology with no one confident or prepared to use it. The good news is that investing in people pays off.
Laying the groundwork
Think of it like building a house: AI and automation are the dazzling appliances and smart home systems - they can do amazing things, but only if the house has a strong foundation and the wiring and plumbing are sound. You wouldn't install a state-of-the-art smart thermostat in a building with a leaking roof and faulty electrical circuits.
For operations leaders and plant managers, the takeaway is clear: before you dive headlong into an AI or automation project, take a hard look at your fundamentals. Are your processes standardised and running as intended? Are your machines reliable, safe, and producing quality outputs with minimal intervention? If not, invest in fixing those basics. Engage your frontline experts. Then, and only then, unleash the power of AI and automation on that solid base.
So, before you invest in the factory of the future, make sure you've perfected the factory of today.
This is a shortened version. Read James's full article on Different Thinking →
If your site needs help turning operational excellence from theory into daily reality, start a conversation — exploratory, no obligation.
Related: Operational Excellence in 2026: A Competitive Necessity for UK Manufacturers · Our services
