In Industry 4.0 manufacturing in Ireland, less than 5% of data is ever used.
At the ISA Ireland Section 4.0 Conference, Ireland’s leading engineers, digital leads and operations managers made it clear why and what needs to change.
If you were in the room on 3 March, you’ll know this wasn’t a day of vendor pitches or blue-sky thinking. Instead, it brought together practitioners speaking honestly about what is working, what is not, and where the real barriers to progress lie.
The message was consistent across every session.
Industry 4.0 digital transformation in manufacturing is not being held back by technology. It is being held back by operating models that haven’t caught up.
Here’s what that means in practice and where to focus next.
1. You Are Sitting on a Gold Mine You Can’t Access
One of the most striking insights from the day was how little data is actually used.
Most manufacturing environments actively use less than 5% of the data they generate. Sensors are running. Historians are logging. Dashboards are updating. However, most of that information remains unstructured and unused.
Data volume isn’t the issue. Data architecture is.
Over the past decade, organisations have focused on collecting and storing data at scale. However, the outputs remain largely retrospective. Dashboards show what happened yesterday. What’s missing is the ability to act on what is happening now or what is about to happen next.
As a result, the focus is shifting. It is no longer about collecting more data. Instead, it is about connecting data and using it to drive decisions across the operation.
Key takeaway: Audit your current data. If you cannot clearly define who uses it, what decision it supports, and what outcome it drives, it is not delivering value.
2. Unknown Variation Is Quietly Driving Your Process Issues
A recurring theme was the impact of unknown process variation.
Unplanned downtime creates problems. However, not knowing why it happened creates bigger ones, because it guarantees repeat issues.
In most environments, unknown variation comes from three areas:
- Factors that are not measured at all
- Factors that are known but not fully controlled
- Interactions between variables that are not consistently visible
For example, one case showed how a simple sensor uncovered a pattern that was previously invisible. Performance degraded in a predictable way before each failure event.
Once teams understood this threshold, they changed their response. Instead of reacting to failures, they intervened earlier and planned maintenance more effectively.
More importantly, when this data was combined with other inputs, such as material variation and environmental conditions, it provided a clearer view of the process. This is where real value emerges.
Key takeaway: Start with one recurring issue. Measure it properly. The insight is often already there, the challenge is making it visible and acting on it.
3. Spaghetti Architecture Is Limiting Your Digital Progress
Many organisations are dealing with what can be described as “spaghetti architecture”, a complex web of point-to-point integrations built over time.
Each new system introduces another connection. As a result, complexity increases. Over time, this creates an environment that is difficult to maintain and even harder to scale.
This creates a predictable outcome:
- Data silos
- Fragile integrations
- Increasing overhead
As a result, the industry is shifting toward more scalable approaches.
Instead of building direct connections between systems, data is published once and made available across the organisation. This allows systems to interact without requiring new integrations each time.
The principles are consistent:
- Connect data across systems
- Add context so data is meaningful
- Ensure governance, security and traceability
In practice, this approach reduces complexity, improves data quality and enables faster scaling across sites.
Key takeaway: Before introducing any new system, consider how it integrates with your existing environment. Interoperability should be a requirement, not an afterthought.
4. AI Without Context Is Just Expensive Guesswork
AI was a major topic throughout the day. However, the discussion remained grounded and practical.
The consensus was clear. AI only delivers value when supported by the right data and context.
To assess where AI fits, it helps to ask:
- Are we informing, recommending, or automating decisions?
- Is this an operational or regulated use case?
- Do we have trusted, structured data to support it?
These questions shift the focus from technology to outcomes.
In practice, one example showed how misleading metrics can be. A model showing 99% accuracy may appear effective. However, in environments where failures are rare, a model that always predicts “no issue” could achieve the same result.
What matters is whether the model identifies the events that actually matter, at a level the operation can realistically respond to.
Therefore, digital literacy becomes critical. Not in building models, but in understanding how to evaluate them.
There was also a clear shift from document-driven to data-driven processes. In document-driven environments, teams manually retrieve and review information. In contrast, data-driven environments allow information to flow continuously, enabling faster and more scalable decisions.
Key takeaway: Do not start with AI. Start with the decision you want to improve. Then assess whether your data is reliable, structured and fit for purpose.
5. Transformation Is a Journey. Start Anyway.
The final sessions brought the conversation back to execution.
Digital transformation often appears complex. In reality, it is a sequence of practical steps built over time.
In simple terms, it involves moving from fragmented or manual processes to a connected environment where data supports decision-making across the organisation.
To succeed, organisations need:
- Clear ownership and accountability
- A defined objective
- A structured understanding of current challenges
- The ability to prioritise and act
Only then should technology enter the conversation.
A key distinction emerged between “proof of concept” and “proof of value.” Too often, initiatives focus on demonstrating technology rather than delivering outcomes.
Focusing on value ensures each step contributes to measurable improvement and builds momentum.
Key takeaway: You do not need a perfect strategy to begin. You need a clear problem, ownership and the willingness to act. Progress comes through execution.
The Bottom Line
Across the conference, a consistent message emerged.
Data is not the issue. How it is structured, connected and used determines its value.
Integration remains one of the biggest challenges. Without a scalable approach, systems cannot deliver long-term impact.
AI will continue to evolve. However, its effectiveness depends entirely on the quality and context of the data behind it.
At the same time, transformation is not a single initiative. It is a continuous process of prioritisation, execution and learning.
Most importantly, the barrier is not technology. It is how organisations operate, make decisions and adapt to change.
As discussed throughout the conference, the role of automation and digital capability is continuing to evolve. It is no longer just about implementing systems, but about enabling real-time decision-making and solving the right problems.
As Industry 4.0 manufacturing in Ireland continues to evolve, so too does the role of automation engineering across pharma environments.
We explore this in more detail in The Future of Automation Engineering in Pharma: Insights from Kieran:
👉 https://lscconnect.com/the-future-of-automation-engineering-in-pharma-insights-from-kieran/