Where do you start when trying to implement the utopia of Industry 4.0?

The world seems to have acknowledged that the Internet of Things will be beneficial in all areas of our lives. Even the manufacturing industry – often described as resistant to change – is beginning to understand that Industry 4.0 is happening and provides many benefits. However, there is still some confusion around certain terminology and exactly what it takes to make a factory ‘smart’.

These are questions that Mitsubishi Electric has sought to answer, building on more than 35 years’ experience of automating factories and implementing smart factory and Industry 4.0 concepts for its customers in the UK.

Mitsubishi Electric marketing and operations group manager, Chris Evans, says: “There is a lot of hot air talked about this subject. The real impetus behind Industry 4.0 comes not just from the link between the plant and the enterprise, [but from the ability] to measure actual performance against an ideal model – the cyber physical system.”

It is this ability for in-depth analysis and continuous improvement that Evans says defines the true spirit of Industry 4.0, but how should it be implemented?

Standards are being developed that will define elements of a system, be they a control or physical elements, as well as communication standards like Reference Architectural Model Industrie (RAMI) 4.0 and IIRA architecture models.

Evans says that once these standards are established manufacturers will have the convenience of a plug-and-play operation.

He explains: “If you go back to when we adopted open network standards, people had something to achieve, we all had a standard to meet and it made it a lot easier to go forward from there.

“If a plant was built from the ground up, using all the technologies that are readily available today, regardless of communication standards, it could be made a smart factory. Unfortunately, the reality is a way off from there.”

The challenge for many manufacturing plants is that their automation systems have evolved over many years, resulting in disparate automation platforms, poor network infrastructure, no data management strategy and very little genuine knowledge of how to get the relevant information out.

“The step change was during 2008-9 when, instead of ripping the machines out when they got old and putting in a new production line, they wanted to make the assets last 15-20 years,” says Evans.

There is also an attitude against data collection and storage in the cloud, with some plant managers wary of storing information away from what they perceive as their domain, despite the proven security that cloud storage offers.

To combat this, Evans says it is important to visualise the plant first and worry about the cloud later. First, he says, define exactly what the manufacturer is trying to achieve, outline what their business drivers are, understand where their problem areas are and what automation currently exists. Then, assess what automation network infrastructure is already in place. Setting expectation levels is also key, as it may take years – and significant investment – to create a fully smart factory. It’s about taking small, but progressive steps.

“The key is to look for quick wins that will demonstrate fast returns against a moderate budget, which prove that the path you’re taking will deliver much bigger benefits over the longer term,” says Evans. “Understanding what is happening at the production plant is essential and infrastructure must be created to achieve this, even if this is approached in stages, whilst always keeping an eye on the end goal. If you don’t know what’s going on, you don’t know what’s going wrong.”

Mitsubishi Electric has undertaken smart factory improvements at its own manufacturing facilities. For example, at its Kani Works switchgear production facility in Japan, upgrading to modular robotic production cells drove a significant increase in productivity and operating rate by reducing the number of stages in the manufacturing process.

Redesigning the production line into a more compact manufacturing cell, utilising robotics and vision as well as conventional automation control, resulted in an 85% reduction in the occupied floor space, which is particularly significant as space is at a premium in manufacturing plants.

Sensors in a cyberphysical system record metrics throughout the process, like the torque on each screw, in real time and model it against the ideal scenario. Evans explains: “This is a true Industry 4.0, smart factory activity: knowing exactly what’s happening on the production floor; referencing what should be happening; analysing the gap and if that is achievable by the control system, automatically. If not, it highlights areas in the process that may need to be changed. There’s your utopia. The whole way along, you’ve got visualisation of what’s happening from the plant up to the enterprise level.”

One of Mitsubishi Electric’s customers posed a problem that wasn’t even considered by Evan’s team. In this particular instance a lot of the plant’s machines were automated and had PLC control but were working in isolation, rather than being networked together, but others weren’t automated at all, including power presses.

The presses are operated by workers pressing left- or right-hand panels. Changing the machine over took 45 minutes. To save time – and thinking they were doing a good job – the operator would press all the panels as one type.

“They explained to us that this was a problem,” says Evans. “In most factories space is at a premium. If, instead of making 200 panels he’s made 300, you’ve got to put them in the store. Now multiply that by all the other machines doing similar jobs.

“That happens because there’s no control, there was no communication from that machine to say how many he’d made. That’s an area we’re focussing on because even if we can find out how many are being made to start with, then by controlling the machine you can stop it. He can’t make any more.”

This production related issue seemingly has little to do with automation but can be solved by automation, according to Evans, but only after having detailed discussions with the correct people in the plant and understanding their goals.

Most plants in the UK don’t have the luxury of being designed from scratch to meet the goals of Industry 4.0, but that doesn’t mean it can’t be done. Evans concludes: “With strategic planning and a structured approach, any plant can reap the benefits of optimised, sustainable, safe production that is energy efficient, all within a fully connected supply chain. The road to digitalisation begins with the first step.”