You can’t move for mentions of AI and how it will change the world these days - but how is it really impacting the manufacturing sector?

We sat our co-founder Theo Saville down recently and picked his brain on the matter - here’s his thoughts on where we are now, and where we are heading.

Q. How is AI transforming the manufacturing sector? Can you highlight some key trends and areas of impact?

Theo: The honest answer is that AI is not currently transforming manufacturing, as the kinds of AI you need to make a genuine difference for the sector haven’t been invented yet.

Manufacturing gets better and faster if you find ways to lower your overheads and costs, and/or find ways to improve your yield and efficiency. But recent developments in AI are mostly in the software realm - image and text generation, and data sifting - and based on large language models. For AI to be useful in manufacturing, it needs to be extremely robust and deterministic, and not built on fuzzy datasets with slightly iffy logic. The kind of funny mistakes that we see from Chat-GPT and other solutions when it comes to answering queries and creating images would, if translated to a factory environment, wreck expensive equipment very quickly.

That said, CloudNC is on the vanguard of developing AI so it can make a genuine impact: our solutions automate CNC machining and programming, and are already making manufacturers more efficient and productive. What we have made is one of the most advanced applications of AI in manufacturing that you can get - we’ve been working on it for 9 years - but, really, it’s only been properly deployed in the market for a few months. So really, we’re right at the very start of finding out how AI can transform manufacturing.

Q. Beyond efficiency gains, what other benefits can AI offer to manufacturers?

Theo: Right now, the benefits are the same as in other industries - AI offers faster access to information, so you can ask questions about complicated subjects and get much faster, accurate answers without having to know and understand all the detail. That’s something particularly applicable to manufacturing, given that previously your way of finding answers out would be to sift through textbooks or find someone with expertise to help.

‍Q: What are some of the biggest challenges hindering the wider adoption of AI in manufacturing?

Theo: The biggest challenge at present is that the solutions available (eg, Chat-GPT) are all geared towards service industries, which makes it difficult for the manufacturing sector to engage, especially if those looking to get involved aren’t especially IT literate.

Q: How can businesses, especially SMEs, overcome these challenges and leverage AI solutions effectively?

Theo: They need to identify tasks that a trained AI can specifically help with. A generic AI is probably not that helpful at this point, unless you are a very large manufacturer with a huge amount of data that you could let it loose on.

Our CAM Assist AI is a good case in point - it’s a solution specifically built for one task, which is to accelerate precision machining through automating programming, thereby solving a big bottleneck in the system. That can save manufacturers an enormous amount of time.

Q: Looking ahead, what exciting possibilities do you see for the future of AI in manufacturing?

Theo: Harnessing AI is a means to address the skills gap by making people multiple more times more productive than they could otherwise be. By taking away the boring repetitive work with AI, you can point your experts towards the areas where their expertise really makes a difference.

Q: How can AI be harnessed to optimize production processes and resource utilization in factories, leading to a more sustainable manufacturing environment?

Theo: Generally it can’t right now, unless you are a CloudNC customer! Solutions built to have an impact in manufacturing are few and far between as it’s a very challenging use case. That said, we will first start to see an impact in simple tasks like picking and placing in warehouses: areas where being very precise isn’t as much of a critical requirement, and if something goes a little bit wrong it’s ok.

In the future, the big difference that AI will make is in enabling greater efficiencies, which will in turn boost reshoring. The reason that manufacturing went abroad in the first place is that labour is cheaper overseas. However, if you can use AI to make machinists at home more productive, then labour stops being such a cost factor, and you can carry out more manufacturing closer to home.

Q: As AI plays a growing role in decision-making within manufacturing, what ethical considerations need to be addressed?

Theo: Broadly speaking, most people don’t want to work in manufacturing, and that’s a big reason why there’s a skills gap. So technology like ours isn’t displacing any jobs. Instead, it’s a tool that makes skilled people and newcomers alike more productive, and making their jobs more interesting and better paid.

Q: What skills and workforce training are needed to ensure a smooth integration of AI into manufacturing operations and empower human workers to collaborate effectively with these intelligent machines?

Theo: One nice aspect of AI is that it makes it easier to carry out complex tasks with a lower skill level. However, that means you need to build digital literacy in the next generation of workers so they can use the tools that companies like ours are building, so they can use them effectively. At school, that might require replacing a few English lessons with software engineering.

‍Q: The UK government has launched various initiatives to promote AI adoption. How can these initiatives be further tailored to support the specific needs of the manufacturing sector?

As a company, CloudNC has been backed by the UK on more than one occasion. The grants programs for advanced technology are helpful, and the UK is a good home for us - it has a strong high-value manufacturing base, great investors and talent, and the right mix of most things.

The only thing it lacks is the size of the manufacturing market, and so as a software company we’ve had to look abroad for sales from day 1 of going commercial.

Q: Can AI play a role in reshoring manufacturing, or will cost and efficiency concerns continue to favour overseas production?

Yes, and CloudNC’s AI specifically helps by making machinists more efficient, making manufacturers in turn more productive. So reshoring follows as a result.

Q:  Imagine a future where "intelligent factories" are commonplace in the UK or US. What would this future look like, and what are the potential benefits and risks to consider?

There is a much greater risk than that of having lots of intelligent factories in the UK or US, which is: what if there are no factories or manufacturing base at all.

A country needs to be self-sufficient in certain regards - defence, food - as otherwise you are vulnerable, especially if you’ve just lost your largest trading partner. The UK and the US need competitive factories to run effectively, and I hope our solutions will help them to run, and run productively.

‍