Artificial intelligence has the potential to transform Ireland’s manufacturing base, and collaboration and knowledge sharing is the key to unlocking the huge potential here, say industry experts.
While adoption of industrial AI here is still in its early stages and there’s no data revealing the level of AI investment, an ecosystem is now developing to encourage it, involving State-sponsored bodies, industry groups, academics and firms.
Companies have adopted a cautious approach, says Domhnall Carroll, chief executive of Digital Manufacturing Ireland, a State-backed body responsible for promoting digital innovation. Carroll says barriers to AI adoption include companies not having a strategic roadmap, lack of financial resources for investment and many not having figured out the HR implications of AI adoption.
“Adoption levels now are quite low. I don’t think that anyone is applying AI in a blanket way to their factories. I think you are seeing testing and people making themselves aware of the technology, and we’re seeing a lot of early-stage pilot projects.”
The payback and the expectations of what AI will deliver are not always clear to companies, he says. AI works best in environments where there are a lot of standardised processes and a lot of data. That can typically be in areas such as electronics manufacturing, for example, but there are many sectors and sub-sectors where deploying AI widely – or in discrete processes – can cut costs and increase quality and service levels.
“Over the long term, many believe it will be hugely valuable but over the short to medium term firms are wondering what projects they should launch and how would they assess their success, so a common challenge with AI in manufacturing is a belief that companies need to see the benefits before they spend money.”
Getting access to diverse data sets that are representative of real-world manufacturing environments helps build more accurate and robust AI models
— Linda Barron, chief executive of the Irish Centre for Business Excellence
Raising knowledge of the potential of AI and sharing best practices are key steps to de-risking initiatives at the firm level, he says.
The Irish Centre for Business Excellence is among the bodies involved in promoting peer collaboration regarding AI among its membership base of high-tech companies.
“Collaboration helps to accelerate the adoption of AI through the pooling of financial, technology and talent resources which might be unattainable individually,” says chief executive Linda Barron. “Data is a critical component for training and improving AI algorithms, for example, and getting access to diverse data sets that are representative of real-world manufacturing environments helps build more accurate and robust AI models.”
Implementing AI technologies in manufacturing can involve risks such as data security, privacy or disruption to existing processes and collaboration with peers allows for collective assessment of these risks, she adds.
The launch last week of the Visual Cognitive Manufacturing Group, supported by the IDA, is the latest initiative in this area. This group, led by several firms in the medical devices and pharmaceutical sectors, including Boston Scientific, Medtronic, Abbott, Johnson and Johnson and West Pharmaceutical Services, will share knowledge in areas such as predictive analysis and machine learning as well as virtual and augmented realities.
[ New taskforce will help pharma and medtech groups use tech to boost manufacturingOpens in new window ]
An international study highlights, however, that many companies are struggling to introduce AI in their production facilities.
The Institute for Learning and Innovation in networks at Karlsruhe University of Applied Science interviewed more than 650 global manufacturers about their AI adoption practices.
The study found that organisational factors, such as digital skills, company size and R&D intensity had the greatest impact on the adoption of AI in manufacturing. Research-intensive, knowledge-based and service-oriented companies were found to be the best adopters and tended to roll out AI technologies not only at their domestic facilities but also at their foreign production sites, it added.
It is no surprise then that the medtech and pharmaceutical industries here are among the most enthusiastic adopters of AI.
Dr Sinéad Keogh, director of the Irish Medtech Association within Ibec, says companies are integrating AI into their operations, as part of a broader digital technology transformation.
“We’re seeing digital technology – including AI – supporting more efficient and more effective manufacturing of products. We’re seeing very high levels of investment by the medtech and biopharma sectors as well as engineering industries in advanced manufacturing technologies. We’re hearing about a lot of companies using machine learning in areas such as predictive maintenance, order tracking and delivery scheduling, for example.”
We need to ensure in Ireland that we are sharing our challenges, we’re sharing what we know about existing and emerging technologies, about vendor performance and what we are seeing globally
— Bill O’Leary of West Pharmaceutical Services
About 350 AI-enabled medical devices entered the market last year in areas such as medical imaging and the treatment of cardiovascular disease, she says, while AI is also being widely used in product development, including by many high potential Irish SMEs working in this sector.
One company that has begun integrating AI into its operations here is West Pharmaceutical Services. Bill O’Leary, director of global operations and manufacturing digitalisation says that its focus is on capturing data from as many of its systems as possible. Among the benefits of this is that it can more easily highlight any anomalies in the production process, for example. “It’s early stage for now. We have not got into predictive mode yet, but we will,” he says.
O’Leary sees the benefits of peer collaboration in this area.
“There are common challenges here that are agnostic to the sector, so it makes sense to share information that is not IP sensitive. We need to ensure in Ireland that we are sharing our challenges, we’re sharing what we know about existing and emerging technologies, about vendor performance and what we are seeing globally. The ecosystem of support and collaboration here is among the best in the world.”
History shows us that better manufacturing results in more manufacturing jobs and better jobs
Ireland’s well-resourced base of FDI companies may have a head-start in this area but SMEs are not precluded from leveraging AI, he believes. Typically, the decision-making process is longer and the validation threshold much higher in the multinational sector than for smaller nimbler operations who can deploy AI technologies much quicker in many cases.
Concern has been expressed that AI adoption raises the prospect of mass lay-offs in the manufacturing sector. Carroll believes this is overblown and that the instinctive knowledge of many high-skilled workers will ensure their places on the production line won’t be replaced by AI-based systems any time soon.
“History shows us that better manufacturing results in more manufacturing jobs and better jobs. There are large parts of industry where there is feedback from the process to the person, not through the machine system but from a person’s observation and contextual understanding.
“These are really hard to replicate and undesirable to replicate because every case is different. The more complex and broader the challenge, the more likely it is that people will have a strong role in that. AI and automation are there to support people, – it’s not the other way around.”
