Four paths to rapid operational value on the factory floor

The performance of a production line has historically been driven by inherent operator knowledge: the deep, intuitive expertise of a veteran workforce. Today, as that expertise nears retirement and labour markets tighten, manufacturing is reaching a pivotal turning point.

Closing the gap between technology potential and operational impact

Despite projections indicating digital transformation in global manufacturing will increase to US$3.62 trillion over the next five years, many manufacturers are still waiting for a return on their existing digital projects. Under pressure to lower costs, the traditional toolbox of operational improvements and reliance on human expertise is delivering diminishing returns. With market opportunities accelerating, businesses are increasingly looking to AI and machine learning to capture the step change in performance that the old approaches can no longer unlock. But too often, these digital solutions are deployed against unclear targets or in isolation from operational goals, creating hesitation on the factory floor rather than the rapid ROI and speed that leaders are seeking.

For supply chain leaders today, the challenge is to utilise emerging digital tools to extract maximum performance value and reduce waste with short ROIs. Based on our experience collaborating with global operators, four paths consistently emerge to drive near-term value at pace on the factory floor.

1. Optimising manufacturing operations by prioritising ‘edge’ entry points

Many manufacturers stall in their transformation journey before they’ve even begun, weighed down by the assumption that autonomous production transformations require massive hardware overhauls. The belief that you must replace the old to innovate for the new is precisely what makes proving ROI so elusive.

Instead, manufacturing leaders should first consider where capacity gains can be realised from assets already on the floor. For example, traditional process assumptions mean most machines run at fixed, conservative speeds, designed to centre line processes. But by using algorithms to identify the real-time sweet spot, you can maximise throughput without crossing the threshold into waste.

This logic extends to the product itself. Dynamic use of AI can control processes to achieve tighter capability windows. This could mean creating recipes that reduce ingredient costs by operating at the most cost-efficient level of acceptability, rather than staying safely on the centre line and over-burdening its Cost of Goods with expensive raw materials. 

In action on the factory floor: We saw this approach deliver tangible impact with a leading global personal care manufacturer. Machine speed setpoints on a filling line had been fixed to suit early run conditions, and once the process stabilised they were rarely revisited because the line appeared to be running well. This created an invisible loss of capacity across the full production run. By using real-time data to adjust speed dynamically, we were able to optimise throughput and deliver a 20 percent increase in output using the same asset.

2. Real-time manufacturing control: Sustaining factory performance beyond big data

For years, big data has been positioned as the foundation of the modern manufacturer. It excels at long-term benchmarking, but is often too disconnected from the immediate needs of the factory floor. To achieve ROI, manufacturers don’t need more data, they need targeted insight they can trust, providing a direct pipeline to improvements they can act upon.

Shifting towards direct data is key: 

• Transition from monitoring what happened to controlling what is happening right now. 
• This should begin with scientific hypothesis-led insight, involving different experts in engineering, AI, and applied science to gather insight that is robust but quick to implement. This will continue to interrogate the factory floor but replace inference with evidence.
• From there, simulations can be run, steering clear of over-engineered digital twins that attempt to model entire systems upfront and often require months to deliver results. Instead, organisations should build quick, lightweight, adaptive models that test ‘what if scenarios’ in days.

This precision also exposes invisible losses. Micro-stoppages under five minutes rarely appear in manual logs and when they do occur, they’re too high level to provide valuable operational insight. Over time, they accumulate into significant, unrecovered capacity loss. Automating their capture reveals the true heartbeat of the line and eliminates the interruptions that undermine high-speed production.

In action on the factory floor: We applied this approach with a global pharmaceutical manufacturer, investigating inconsistent quality across identical international lines. Armed with scientific data, we uncovered second-order relationships, including humidity effects that surfaced only intermittently. By defining an ideal, golden batch for each operating environment, the manufacturer stabilised output across all sites and significantly increased operational yields.

3. The self-funding roadmap: Proving value before scale

A heavy upfront capex gamble is unlikely to deliver the desired short-term returns, but has the advantage of big, long-term paybacks. Adopting a self-funding roadmap, a strategy that uses immediate savings from early optimisations to fund the next stage of the journey, could unlock the larger paybacks, while still accessing the short-term returns.

In action on the factory floor: We have seen this work in practice with multinational manufacturers running multiple optimisation programmes in parallel, each with different timescales and returns. In one global consumer packaged goods business, early savings generated in one product category were deliberately pooled and reinvested to fund improvements across others.

This approach also changes how investment decisions are made. Rather than leading with the most ambitious initiatives, leaders should prioritise opportunities based on:

• Speed to value
• Ease of implementation
• Capital efficiency.

With the global consumer packaged goods business above, the result was a series of pragmatic improvements that delivered measurable returns quickly and justified further investment. These could be as simple as reducing the trim waste of packaging, lowering energy modes for dormant machines, or upgrading machine parts to more than triple the mean time between failures on a labelling machine, for example.

Importantly, this financial journey must be supported by an organisational shift that ensures test sites aren’t penalised on their own P&L for hosting pilots. While value is generated on the factory floor, risk must be managed at the network level to remove the structural disincentives that cause plant managers to block new initiatives.

4. Closing the adoption gap: Codifying expertise

When your most experienced operators walk out the door, they take the factory’s operating manual with them. To secure the future, leaders must switch undocumented instinct and institutional knowledge into demystified processes, settings, and setpoints. That shift demands collaboration across the shop floor, underpinned by data. On most lines, operators each run the same equipment differently, and while performance may appear acceptable, this variability is the enemy of reproducibility and line predictability.

In action on the factory floor: We saw this challenge first-hand when we were brought in as an independent party by a major global machine builder facing potential contract loss with a multinational drinks manufacturer. Across multiple sites, the client believed the equipment was underperforming. After a short period of time on the shop floor capturing data and observing behaviours, we found the machines were not the issue; it was how the lines were run.

The solution was twofold. First, we brought operators together to surface differences in approach and share knowledge. Second, we removed unnecessary variability by locking down parameters both digitally and physically, ensuring operators retained control where judgement mattered.

Resistance is inevitable, which is why change must be proven. When experienced operators see measurable improvements in production yield or waste reduction, creating greater ease, and operator efficiency, scepticism gives way to adoption. Building teams, structures, and processes to share hard-won intuition creates a critical window to capture expertise before it disappears from the business.

Engineering operational excellence in an era of constraints

Manufacturing leaders who target their digital resources towards codifying human insight, faster ROI, and value-led interventions will outmanoeuvre the industry’s operational squeeze. Moving beyond the traditional metrics will unlock invisible capacity to ensure the best performance is achieved every single shift.