In order to make the most of Industry 4.0 and what is has to offer manufacturers, businesses need to know how to go about using the technologies correctly. This is the message from industry experts who attended the Industrial Internet 4.0 Summit at the University Technology Sydney (UTS) in late February.
UTS professor of engineering John Broadbent said manufacturers need to ensure their products are open sourced so they can easily connect with other technologies and equipment.
“Manufacturers, particularly with their machinery, are being asked to make sure that the kit they are buying is open source. There’s no point buying a piece of equipment that in five years’ time, or even tomorrow when you get the gear, you can’t get data out of that equipment to then put up into any system,” said Broadbent.
Where people have used old systems and are for example moving to the cloud, there is a worry that if it is not open sourced or easily accessible and transferable, that information will be lost.
“What I’m seeing that’s a concern to me is that we used to have site-based silos of information. Getting data out of those silos has been part of a challenge,” he said.
“One of the concerns I see in the market place is that we’ve effectively in some cases done nothing more than uplifted this site base on premise silo into a cloud-based silo. Beware that if you’re going to move data out of the edge coal face and you’re going to put that into the cloud – who owns the data?
“Do you have access to that data and can you share that data amongst other systems?”
When investing in new technologies, Broadbent suggested that manufacturers think carefully about what they are investing rather than trying to jump from nothing to the best without doing research.
“One of the things I’m concerned about is that as organisations start to invest in this technology very few are not risk adverse. “I’ve seen a lot of organisations that just don’t know how to get from Industry 3.0 to 4.0.
“The two precursor steps that are needed to start the journey are computerisation and connectivity.”
Broadbent explained that without the equipment to acquire all the data in a company, and without the right infrastructure in place such as cyber security and network switches, “you’re stuffed”.
“You’ve got to start somewhere and the way to start is through small proof of concepts. Set them up for success, not for failure.”
There needs to be governance and policies in place with a medium-to long-term plan, said Broadbent.
“I had a customer recently in a fully integrated processing plant that was built four years ago. The engineer in that plant went out and bought a $15 million purpose-built lasagne line. He doesn’t own the source code in the chiller, he doesn’t own the code in the oven. The stuff is password protected, he can’t get in and change the IP address to put it in his network.
“I look at that and think that’s just insanity,” said Broadbent.
While being able to access the information in the first place is pertinent, manufacturers can often struggle to use the data they receive.
CSIRO research program director Dr Leon Prentice explained that knowing what to do with machines and the information they can provide is important.
“There are two parts to Industry 4.0 – gathering the information you need and doing something useful with it. Where people break down is generally on one of those two.
“Either they are gathering useful things, or not gathering things, or they are not actually doing anything with it, but just sticking it into a giant excel file that just keeps growing and they don’t have the tools to do anything with it.”
Prentice said, by knowing how to use and read information, manufacturers could save on simple aspects of their factories such as knowing when to turn off the lights based on the amount of usage.
UTS professor Michael Blumenstein said people need to be aware that there is no single solution to Industry 4.0.
“There is no general Artificial Intelligence (AI) technology. “There is not possibility of having one platform that can solve all of your AI needs.
“Usually they are tailored solutions. When you design an AI system or buy one there is no sure bullet.
“Be aware of any one that tells you, ‘I’ve got the solution for you’,” said Blumenstein.
While there are some issues to be aware of, Blumenstein said that manufacturing is generally the perfect area to apply AI.
“The AI can also be used for prediction ahead of time and this is where it will be a really powerful thing for people to use.
“The AI can give you an indicator when something is going wrong in real time. But where I think the real power of the AI is, is in the prediction. This is extremely important.”
Blumenstein gave an example of a plant not being monitored properly. If there’s no in-depth preventative maintenance plan and no ongoing inspection process, the plant can explode, costing millions of dollars, he said.
“You’ve got two ways to look at things there’s reactive maintenance, which is the old-fashioned approach – there’s obviously manual inspection involved – and there’s the other part, which is preventative maintenance. These days, a lot of that preventative maintenance can occur with AI, with robotics and other automated approaches.
“What that can do for any organisation is save millions when things go wrong,” said Blumenstein.
However, with any system that collects large amounts of data, there can be issues that also need to be prevented where possible. “The challenge of course for anyone is you need a lot of data for this. A lot of the learning algorithms are very data hungry and if you don’t have the data you can’t train your technology as effectively. That’s why it’s so important that the data is so readily available,” he said.
“It needs to potentially be stored and looked after to be readily available.”
The Industrial Internet 4.0 summits was held from February 26-27, with senior level executives from the Australian manufacturing industry gathering to discuss developing their digital strategies.
The summit attracts leading authorities from government, business, academia and industry associations keen to share their experiences and to look at best practices for manufacturers keen to implement change across their organisation.