AI investments to transform manufacturing in 2025?

2024 was widely regarded as a mixed year for global manufacturing. While remarkable innovations, projects, and products were once again numerous, there were still a number of persisting macro-economic challenges for businesses to overcome. The predictions for 2025 are optimistic, evidenced in Deloitte’s 2025 Manufacturing Industry Outlook. These predictions are routed in the movements, aspirations and investments of manufacturers. Predictions that go “well past hopes,” according to a bullish Tim Gaus, Deloitte’s General Manager for Smart Manufacturing.    

Moving the needle 

Generative AI investments are predicted to increase this year. A technology that is often spoken about with excitement, and not just by solution providers trying to sell their AI platform. Undoubtedly, manufacturing businesses are recognising the benefits. That said, adoption last year was slow – companies are still testing the water.

“We recently ran a survey across industry, and we found that over 80% of companies have had some degree of activity in gen AI,” said Tim.

“We have seen some pilots and proof of concepts, but a relatively small fraction of businesses have actually scaled these solutions across their network. We see that dramatically shifting this year.”

He continued, “There has been a lot of talk around the future of agentic generative AI, we believe this is going to be very important to success in the manufacturing domain, and it’s one of the keys areas where we see folks moving from pilots into truly scaled solutions.”

But while manufacturers see the benefits in certain technologies, adoption can often be slow. For whatever reason; be it budgetary constraints, or fear of failure, companies constantly find themselves in pilot purgatory. Why would it be different with gen AI projects?

Tim answered, “One, you don’t need to be perfect to get going. Gen AI is one of those release valves that allows you to retain knowledge and help your workforce in a very meaningful way without having to have all of your data in a perfect spot.

He continued, “The second reason is that Gen AI can get you to a value that is not necessarily huge in monetary terms, but meaningful to your people, and instrumental to moving the needle.

“The two hours that spent creating that report in your office are now two more hours that you can spend on the shop floor. Gen AI can create that report in just a few minutes.”

Tim’s point here, is that small wins can achieve big traction. But when we talk about moving the needle on gen AI investments, it is big wins that the sector is targeting as a result – that’s the promise that AI is dangling in front of us.

In this context, we are not talking about routine applications. Deloitte’s prediction is that we will continue to see gen AI used in predictive maintenance and repair, but it will begin to move to more complex and robust concepts, like product development and design.

“The problem solving capabilities of gen AI are starting to get very interesting,” said Tim.

“When you start putting more analytical questions against it, and when this is properly engineered, it can come up with very robust answers. People on shop floors just don’t have the time to do this sort of analysis. Generative AI unlocks that time and actually allows you to solve a lot more complex problems and get to true root cause.”

He added, “I think we are going to see a number of unlocks here, by applying this new tool to our toolkit, using some of the classic methodologies to find answers a lot more quickly than we have done in the past.”

Tech wellness

A commonly touched on point, when integrating and scaling any new solution or technology, is the need for solid foundations, tech wellness as it is referred to here.

“It’s critical,” said Tim. “There are some lower hanging fruits and quick wins that you can achieve with some of these modern techniques, but the true unlock comes from industrialised, fully scaled solutions, where you need to have a data model that is contextualised and scalable.”

He continued, “This enables the ability to move these solutions across your lines, across your facilities, in a seamless fashion. And the core to this is industrial data and your manufacturing execution systems.”

The survey that was run by Deloitte was mentioned earlier. Of the manufacturers that gave responses, a large portion admitted to having not fully embraced MES. While investments in new technologies are expected, there is also a predicted, and a needed investment in core and the foundational tech, because this is critical for long term success.

The prediction is that we will see a more cohesive technology strategy in 2025. One that will strive for tech wellness, but will be also drive value from more advanced techniques.

“In the past, companies would say they have to get the foundation 100 per cent in place before even talking about more advanced stuff,” said Tim

“People are recognising that they will be left behind if they wait that long. And some might argue that getting the foundation perfect is an impossible task. If not impossible, definitely not a quick task.

“I think the strategy will be a hybrid one, where you are using the investments you have made already and leveraging as much advanced outcomes and value as you can, while still investing in the core and driving the overall foundational aspects forward.”

Cracking the code

Outlooks for the business year ahead are numerous; they come from businesses, bodies and individuals right across the manufacturing ecosystem. On occasion, they are aspirational, based on ideals and hopes. Or, to put it more kindly, educated guesses on the direction of travel. Deloitte’s 2025 Manufacturing Industry Outlook, Tim assures us, is based on evidence.

“I would put them well past hopes,” he said. “Although I absolutely do hope that they happen, because I think this is going to move our manufacturing base forward substantially when they come to bear.”

“These are things that you’re already seeing from your customer base?” I asked.

“These are concepts that are being built on fast,” said Tim. “We have folks in the manufacturing space who spent a bunch of money trying to move their company forward and understand where that value would be. Now, they’re starting to see where that value lies.”

He concluded, “To me, this is the big setting spot where companies who have been on this journey for a while are now really starting to crack the code and get themselves moving forward.

“I absolutely hope that our manufacturing base will reach a brand new era of productivity, but I don’t think these predictions are hopes themselves. I think they’re actually an extension of very promising past work.”