With a fresh 10-year blueprint for economic growth, the UK’s Industrial Strategy puts advanced manufacturing at the heart of its ambitions. James Watson, Partner, and Rachel Noll, Manager for Data & AI at IRIS at Argon & Co, explore how smarter use of data, automation, and robotics can help manufacturers unlock productivity, resilience, and long-term growth.
Credit: Shutterstock / chrisdorney
The UK Government’s newly launched Industrial Strategy was long in the making, but has arrived with bold ambitions. Its 10-year roadmap for economic growth has a firm bet on advanced manufacturing as one of the eight high-potential industries in the UK, along with sectors like financial services, clean energy, and life sciences.
For many operating in this sector, this support couldn’t have arrived soon enough. Manufacturing has been pushed from disruption to disruption, hampered by inflation, persistent labour shortages, and global supply chain crises. Businesses have been urgently calling for tools to help them do more with less, and, against this backdrop, the Government’s commitment to invest in digital transformation and skills has been widely welcomed.
The Industrial Strategy features investment in specialist advisory services and organisations to increase technology and robotics adoption across advanced manufacturing. But the big question is now whether it will deliver the change that manufacturers are hankering for, especially in relation to smart manufacturing.
How manufacturers can get smart: in five stages
Central to the Advanced Manufacturing Sector Plan is a push to scale the adoption of robotics, data, and advanced digital technologies. While cutting-edge automation and predictive AI are becoming more accessible, many manufacturers – particularly SMEs – still lack the maturity or infrastructure to implement them.
The Industrial Strategy aims to bridge this gap, announcing a new Robotics and Autonomous Systems (RAS) programme, backed by an initial investment of £40 million. This will establish a new network of Robotics Adoption Hubs – physical centres with the expertise, equipment, and connections to accelerate firms’ adoption of robotics. These will be designed as a ‘one-stop shop’ to help end-users invest in RAS technologies in a safe, low-risk environment.
However, smarter manufacturing also needs to be backed by operational visibility and a strong data foundation. Here’s how manufacturers can embark on this journey successfully:
Stage one: Increase operational visibility
Manufacturers first need sight of their core operational metrics to define and monitor performance. After all, you cannot improve what you don’t measure.
Many manufacturers still rely on paper-based reports and inconsistent metrics, making it hard to compare shifts or pinpoint problems. Without operational visibility, actions tend to be reactive and retrospective. Perhaps a shift has underperformed, but without reliable data, it’s impossible to identify the cause.
The first step is defining consistent metrics across all shifts – such as operatives per line, output per line, downtime reasons, or quality defects. Even simple tools like whiteboards or spreadsheets can instil the habit of consistent data capture and begin building a mindset of continuous improvement. The input might be manual and prone to human error, but it provides a common point of reference and highlights areas needing further insight.
Stage two: Build deeper operational insight
Capturing data in an automated format is inherently more reliable, as it doesn’t require human interpretation. Data such as scan times, equipment health and performance, and employee clock-in and out times can feed into visualisation tools like Power BI or Grafana, helping to spot trends and anomalies over time.
Data is ideally stored in a data warehouse to allow for secure deposit and retrieval in a structured format. Layering information from different sources can reveal patterns. For example, does the mechanical equipment perform consistently at all hours? Are reworks linked to break times?
Organisations may spend longer in this phase retrieving, cleansing, and analysing data, but it’s a vital foundation for future analytics.
Stage three: Apply predictive analytics
One of the defining features of smarter manufacturing is being able to predict what’s happening next and act on it – and predictive analytics can bring this to the factory floor. With knowledge of trends, organisations can begin to form corrective courses of action, strategies of intervention, and avoid downtime. For instance, if the data shows that breakdowns spike after 100 hours of runtime, repairs and servicing can be scheduled in advance. Or, if absenteeism spikes after bank holidays, extra staff can be rostered.
Stage four: Use prescriptive analytics
At this stage, it is assumed the organisation has a strong data foundation. Prescriptive analytics recommends specific actions based on historical feedback loops: detecting a trend, initiating a response, and measuring its effectiveness.
By combining data sources, like weather, complaints, and inbound profiles, organisations can run probability-based models to suggest specific checks or actions. However, human judgment is still required to execute or validate these suggestions. To build trust, models should offer tracing to help users understand why a decision has been made.
Stage five: Become self-optimising
At this final stage, responses are automated, based on high confidence in the data and models. Trust in data is key to achieving full insights maturity. Getting here has likely taken time, learning, and refinement, and as a result, can be relied upon with little human intervention. Like Google Maps rerouting you in real-time around traffic, self-optimising systems react instantly to disruptions – the user only needs to accept or decline the suggestion.
A “human-in-the-loop” retains a level of control, but decisions can be made in seconds. While full automation across the value chain is ambitious, it can be prioritised in high-value areas.
The human factor
While the Industrial Strategy is welcomed with open arms by most in the industry, success still depends on people as much as policy. While the journey is data-driven, people are the linchpin to progress – or the lack of.
Resistance to change is common. Humans simply cannot process large volumes of data as effectively as a machine can, but their insight is vital for interpreting results and providing context. Ultimately, the most effective smart manufacturing journeys have a perfect blend of human intuition with machine intelligence.