The Next Evolution of Manufacturing ERP: From Data to Context

In common with just about every other sector of the digital economy, manufacturing is now rapidly moving from AI assessment and experimentation to deeper real-world adoption.

A quick look at the numbers reveals where things stand. According to one study published late last year, for example, the UK is leading Europe in the use of machine learning technologies on the factory floor. Specifically, 53% of domestic manufacturers are already using AI, with 98% overall planning to implement it. As the report authors point out, “AI is no longer just an emerging trend for UK manufacturers, it has become the driving force behind their transformation.” 

If adoption and implementation rates are this consistent across the industry, the question then becomes where meaningful differentiation will emerge. At present, manufacturers tend to focus on AI for specific use cases, such as predictive maintenance, quality control, supply chain management, and product design. Impressive as this is, for the next phase of development, AI must be used to interpret and act within this operational context rather than analysing data in isolation. 

As a result, manufacturers are likely to prioritise the deeper integration of AI with existing Enterprise Resource Planning (ERP) systems, which have long served as the operational backbone by coordinating processes such as procurement, production planning, inventory management and logistics, among many others. These systems capture the relationships and constraints that govern day-to-day operations, and as such, they provide a ready-made structure that enables AI to interpret signals and support decision-making. 

Turning data into action  

In this context, one commodity manufacturers are now rarely short of is data. Their systems generate massive volumes of information at each and every stage of the process. As any complex data-intensive organisation knows, the real challenge is not collecting data in the first place; it’s that this information is often fragmented across different systems, making it difficult to build a complete picture of what is happening. 

For many manufacturers, this typically means relying on retrospective reporting or manual interpretation, which slows decision-making and limits the value AI can deliver. When AI is integrated with ERP data, however, it can interpret how procurement, production, inventory, logistics and finance interact in real time, and by doing so, fully extend intelligence into operational processes.  

For example, a production delay caused by a late material delivery rarely impacts the shop floor alone. It can also influence labour allocation, downstream scheduling, customer delivery commitments and financial forecasts, among other key processes. But when AI operates on ERP data, it can identify these knock-on effects early, helping managers make adjustments before disruption spreads more widely and, in many cases, unnecessarily. 

This is far from aspirational, with a study by McKinsey suggesting that AI-driven predictive maintenance alone can reduce machine downtime by up to 50% and increase productivity by 20–25%. 

The integration dividend

With these issues in mind, what will the future look like? Firstly, the deeper integration of AI tools into enterprise processes will enable manufacturers to move toward a more adaptive operational model.  

In practical terms, instead of relying on periodic reporting cycles as is generally the case now, enterprise systems will increasingly monitor operational conditions consistently. For example, AI agents will be able to observe performance signals in a true end-to-end format and, armed with that information, adjust relevant processes as required. 

Over time, these systems will also begin to incorporate historical operational patterns, such as seasonal demand fluctuations or regular maintenance cycles. Another important benefit will be the ability to capture the tacit knowledge that experienced members of staff routinely apply as part of their decision-making processes. For example, experienced planners often know which suppliers are likely to deliver late or which production lines slow down under particular conditions, allowing them to adjust schedules before problems arise. At present, this kind of insight is largely invisible to AI systems. 

But when embedded within enterprise systems, this knowledge becomes accessible at scale, enabling organisations to apply consistent operational logic across complex environments. The result is a gradual shift away from static automation toward more adaptive manufacturing operations, where systems continuously interpret operational conditions and guide decisions. 

As for the impact of this innovation on the existing manufacturing workforce, this kind of automation isn’t going to arrive at the expense of human labour. Indeed, research indicates that instead of replacing workers, “UK manufacturers are using AI to enhance roles and address labour shortages. Thirty-eight per cent of UK companies plan to upskill existing talent, up from 30% last year and well ahead of the European average of 30%.” 

It is expected to help deliver a win-win for manufacturers, especially in an environment where the UK has approaching 50,000 vacancies, with 36% seen as hard to fill because applicants lack the appropriate skills, qualifications or experience. 

The real promise of AI in manufacturing, therefore, will not come from isolated applications, but from how effectively it is embedded within an enterprise’s operations. When integrated with ERP systems that capture the complexity of the entire environment, AI can move beyond analysis to support faster, more informed decisions, and manufacturers will be better positioned to build more adaptive, resilient operations.