Insights from ManuPharma 2027 industry research

Over the past few months, we've been speaking with manufacturing and technical operations leaders across pharma as part of our research for ManuPharma 2027.

We wanted to understand what's actually sitting at the top of the manufacturing agenda – beyond the headlines and technology hype.

The conversations covered everything from AI and automation to batch release, ageing assets, workforce challenges and the future of European manufacturing.

Some of the answers were what you'd expect. AI came up a lot. So did automation, digitalisation and Operational Excellence. But the more interesting part has been what sits underneath those big themes.

People aren't particularly interested in talking about AI for the sake of AI. They want to know where it's actually useful. Manufacturers have plenty of data, but they're still working out how to get more value from it. Companies are investing in automation, but that creates its own problems around integration, validation, skills and adoption.

And some of the biggest challenges we heard about weren't particularly futuristic at all – things like slow batch release, ageing equipment and finding enough downtime to modernise existing facilities.

Here are the issues that came through most strongly.

AI – but with a much more practical focus

AI was probably the least surprising theme to emerge. What was more interesting was how people talked about it.

The conversations weren't really about autonomous factories or replacing entire manufacturing teams. They were much more practical.

Could AI help identify the cause of a manufacturing problem faster? Could it analyse production data for predictive maintenance? Could an operator ask an AI assistant a question about an SOP rather than spending time searching through documents?

One person we spoke to described the ambition as getting to "real AI" – actually using the data manufacturers have been collecting rather than just continuing to collect more of it. Another suggested AI could act almost like "first aid" for operators, giving them quick access to the information they need on the shop floor.

That feels like an important shift. The question isn't really "Should pharma use AI?" anymore. It's "Where is it useful enough to justify implementing it?"

The systems underneath AI might be the bigger problem

This came up several times. It's difficult to do much with AI if the information you need sits across ERP, LIMS, QMS, OT systems and various legacy platforms that don't easily talk to one another.

One of the people we spoke to is working on a large ERP transformation where Quality processes span ERP, LIMS and QMS. The challenge isn't just connecting everything. It's doing that while maintaining the validated environment pharma requires.

Another conversation highlighted how fragmented sponsor–CMO collaboration can still be: SharePoint for batch records, a CMO's own system for deviations and email for planning.

It made us wonder whether pharma sometimes jumps too quickly to the AI conversation. There is still quite a lot of work to do simply connecting the manufacturing environment properly.

Manufacturers have plenty of data. Now they want to use it.

This was another recurring theme.

Manufacturers have spent years putting systems in place to collect more information from their operations. Now the question is what to do with it.

One manufacturer we spoke to is specifically looking at how AI can help analyse manufacturing data for things like predictive maintenance and root-cause analysis.

That's quite different from digitalisation being measured by how many processes have been moved away from paper. The conversation is becoming much more outcome focused.

Did it reduce downtime? Did it help us understand why something failed? Did it remove manual work? Did it help somebody make a better decision?

That's probably a much more useful way of measuring digital maturity.

Batch release came up more than we expected

This was one of the most interesting findings from the research.

We heard about batch release times running beyond 50 days and the knock-on effect that can have across manufacturing and logistics.

In one example, an organisation was dealing with an 18-month backlog. They responded by increasing QC capacity, hiring 76 analysts in four months, expanding laboratory infrastructure and changing the way testing teams were organised.

Release time came down from 55 days to 20 days. It's a great example because there isn't a flashy new technology at the centre of the story. They had a bottleneck. They redesigned the operation around it. Performance improved.

It also raises an interesting question about how manufacturers measure efficiency. There's limited value in producing a batch faster if it then spends weeks waiting to be released.

Operational Excellence is still very much alive

With so much attention on AI and digital manufacturing, you could be forgiven for thinking traditional Operational Excellence had taken a back seat. Our research suggests otherwise.

Cost of Goods, capacity, productivity, waste and reliability are still very real priorities.

One manufacturer we spoke to, for example, is thinking about the capacity it will need over the next five to ten years while also looking at automation and OpEx to reduce Cost of Goods. What's changing is the toolkit. OpEx now sits alongside automation, analytics, simulation and AI.

So rather than digital transformation replacing Operational Excellence, we're probably going to see the two become increasingly difficult to separate.

Automation creates problems after installation too

One comment from our research particularly stuck with us.

A manufacturer said they wanted to learn from peers about managing the productivity dip after new equipment is installed.

It's such a practical problem, but not one you hear discussed at conferences very often.

You can spend months selecting equipment, building the business case, installing it and validating it. Then it goes live.

Operators are learning. Processes have changed. Unexpected issues appear. And suddenly the productivity improvement in the original business case looks a little further away.

Those are exactly the sorts of challenges where hearing from another manufacturer who's already been through it is probably more useful than hearing another presentation about the benefits of automation.

Ageing facilities are creating some difficult decisions

Another person we spoke to described their challenge as a triangle: high asset utilisation, ageing equipment and CAPEX pressure.

It's easy to see the problem. If equipment is highly utilised, there's very little time available to refurbish it. But if it's ageing, you can't keep delaying investment forever.

And when you finally make the CAPEX request, you're competing with greenfield facilities, acquisitions and other investments elsewhere in the organisation.

One interesting point from that conversation was the importance of "storytelling" around the business case. Engineering teams might understand perfectly well why an asset needs replacing. The challenge is explaining the risk, return and operational impact clearly enough to win investment at board level.

That's probably a skill that will become increasingly important as manufacturers try to get more life and performance from existing sites.

People are becoming a much bigger part of the technology conversation

Nearly every conversation about AI or automation eventually leads to people.

What happens to operators? What skills will manufacturing teams need? Do you retrain the people you already have or recruit new capabilities? And how do you introduce technology without making employees feel that transformation is something being done to them?

One person we spoke to felt strongly that employee fear and resistance needed much more attention when companies introduce new technologies or launch new facilities.

Another raised a really interesting question around AI: is the objective productivity or growth? Do companies use AI to produce the same amount with fewer people, or does it allow the organisation to grow and create different kinds of roles?

There isn't a simple answer, but it's a conversation pharma manufacturing probably needs to have sooner rather than later.

Where we manufacture is back on the agenda

Geopolitics came through much more strongly than we expected.

Several conversations touched on the pressure to strengthen manufacturing in Europe, reduce dependency on certain regions and think differently about resilience.

One participant suggested looking at European manufacturing investment, subsidies and government support. Another raised the Western world's reliance on overseas production for critical medicines such as antibiotics.

It makes manufacturing-location decisions considerably more complicated.

Cost matters, obviously. But so do resilience, government policy, sustainability, access to skills and security of supply.

The cheapest manufacturing network on paper isn't necessarily the most resilient one.

And finally: people really don't want more sales pitches

This might be the clearest message we've had.

When we ask people what they want from an industry event, they keep coming back to real examples.

They want to hear from companies that have actually implemented something. What worked? What didn't? How difficult was validation? What happened during implementation? What result did you get? And what would you do differently next time?

Several people specifically said that if a technology provider is presenting, they'd much rather see them co-present with the pharma company that implemented the solution.

One participant was even more straightforward: focus on peer-to-peer learning and avoid vendor sales pitches. Fair enough.

So, what does all of this tell us?

If we had to summarise what we've heard so far, it would probably be this:

Pharma manufacturing isn't short of transformation ideas. The challenge now is making them work in the real world. AI needs the right data and systems underneath it. Automation needs to deliver after go-live. Digitalisation needs to improve something measurable. Old assets still need investment. New facilities still need people. And manufacturing strategies that made sense five years ago may need reconsidering in a much less predictable geopolitical environment.

Those are the conversations we're using to shape ManuPharma 2027, taking place in Berlin on 24–25 February 2027.

Rather than trying to predict what the factory will look like ten years from now, we want to spend more time hearing from the people who are actually changing it today.

Register for ManuPharma 2027 here.