AI to form backbone of Industry 4.0

Posted on 15 Nov 2017 by Michael Cruickshank

Among the multitude of new technologies being showcased at Smart Factory Expo, Artificial Intelligence (AI) for manufacturing is one of the most game-changing.

Graham Cox from IBM explained how Watson can benefit manufacturers. Image courtesy of The Manufacturer
Graham Cox from IBM explained how Watson can benefit manufacturers – image courtesy of The Manufacturer.

Forming the backbone of many of the automated and connected systems critical to Industry 4.0 technologies, AI and machine learning are developments which cannot be overlooked.

At The Manufacturer Smart Factory Expo, many exhibiting companies are showing off their AI and machine learning tech and novel ways of implementing it.

Among these is IBM’s Watson AI platform, which uses machine learning in order to power a vast range of manufacturing tasks.

Watson’s primary function is analyse huge data sets and detect patterns and efficiencies within this data set.

“The first step is usually in our experience is about governing the data and getting a coherent data set […] once you’ve got that you can start to visualise patterns within that,” said Graham Cox from IBM.

In a demonstration, the company showed-off Watson’s ability to discern defective products on a moving production line and to listen for the sound of a machine acting abnormally, so as to preempt potential problems.

The Watson AI even has the ability to monitor workers through sensor embedded clothes or safety equipment in order to help identify OH&S risks in the factory and prevent them from happening in the future.

“We don’t see this as replacing the expertise of engineering and production operations, it’s more about augmenting this capability with tools to help do their job even better,” Cox explains.

Enterprise Resource Planning (ERP) company Columbus is another exhibitor at the Expo, also making use of AI and machine learning to make sense of big data.

Among the services they are offer are smart monitoring of equipment, extending its lifetime through what they call ‘predictive maintenance’.

“Historically we would have had manufacturers who say that ‘We know that widget A will need to be replaced every 6 months’ […] But what they have found is that you have taken widget A off and actually there was nothing wrong with it. Using IoT and predictive maintenance we can say actually how many cycles this has done through thousands of sensors and putting this into artificial intelligence,” explained Julian Smith, a business development manager from Columbus.

The company also empowers manufacturers to analyse huge sets of data – for example social media feeds – in order to help predict what customers want, and then allow manufacturers to beat their competitors to the market with this desired product.