“One of the most fascinating things about digital transformation is that it moves instantly – and slowly – all at the same time. The trends we see on the horizon for Industry 4.0 in 2020 are much the same as trends that we’ve seen growing over the last few years,” says Nigel Thomas, head of manufacturing, A&D, automotive and life sciences at Capgemini.
“The difference isn’t so much in the technology, it’s in the number of companies beginning to utilise it, and the reasons why.”
The big question for 2020, he advises, is which technologies will warrant the investments required and thereby accelerate the UK towards Industry 4.0. Excited as manufacturers are by new technology, their interest is, we know, linked intrinsically to the notion of scalability, efficiency and cost effectiveness. Only then can new products and ways of working really become embedded in the manufacturing system.
The benefits of agility and responsiveness are vital to selling the notion of a connected world and these depend upon the sophistication and adaptability of one’s control solutions and their ability to harness data and feedback at speed. But, says Thomas, there is a very practical consideration that underpins the expansion of 4.0 adoption.
The difference isn’t so much in the technology, it’s in the number of companies beginning to utilise it, and the reasons why
Nigel Thomas, head of manufacturing, A&D, automotive/life sciences, Capgemini
“According to the Office of National Statistics, productivity in the UK fell at its fastest annual pace in five years in 2019. Conscious of this, manufacturers will look to adopt latest technologies like IIoT and automation to improve levels of productivity in the UK,” adds Thomas.
To enable this, he predicts 2020 will see a discernible rise in four aspects as manufacturers seek to boost competitiveness at home and abroad by improved process control.
These include cobots, automation, AV/VR and enterprise applications with AI. Repetitive, low-value tasks will be handled increasingly by software bots and machine learning, with machines becoming more ‘intelligent’ as they learn tasks along the way.
And there is a knock-on effect for workforce skills development. COPA-DATA’s zenon 8.10 software, unveiled at the recent SmartFactory Expo, illustrates how software is being coded to meet the requirements of automation, improving its data processing capabilities as the number of variables in factories grows.
The company’s last such update increased the number of potential value changes from 400 to 150,000 per second and, vital for uptake, it did so without impacting runtime performance.
In the case of AR and VR, we are finally seeing their deployment on the shop floor, says Thomas. “Manufacturers have seen the value of these technologies in the simulation of the reconfiguring of factory floors or production lines and the simulation of working tasks. Embedding AR/VR into manufacturing speeds up workforce training, ensuring more time is spent on the shop floor leading to increased productivity in a shorter time-span.”
Lastly, the ‘megatrend’ of the next five years, artificial intelligence embedded in most enterprise applications – “With computer vision embedded in the machines, manufacturers will be able to understand the working models of machinery and preferred setting for different workloads.”
The 5G era will significantly raise many businesses’ game by hugely increasing the number of sensors that they can connect to equipment, providing massive data opportunities around equipment performance, adds Capgemini.
Business-effective data analysis will depend on the ability of feedback controllers to respond to information in order to regulate the actions of devices and regulate their behaviour speedily.
But Jason Zyglis, VP sales, projects and service for AMETEK Surface Vision [pictured], suggests too many companies are missing the potential of discarded data. The cameras on a system provided by a surface inspection company such as AMETEK might generate tens of gigabits of data for each unit of production – high-resolution raw/normalised image data. Powerful classification tools/technologies help make this data more valuable in regard to taking the appropriate corrective action.
Explains Zyglis: “Current system strategies filter non-critical data in order to provide display of only the most critical information needed to identify the relevant defects. This means a lot of potentially useful data is lost – it may not be about defects, but it could provide the customer with important information about any developing issues with the product or equipment.”
Inspection of defects on the surfaces of machinery and other products or the ground presents particular challenges when aligning control processes for Industry 4.0, outlines Zyglis. “We create images and defect maps, and we understand where things are, but we don’t necessarily have the feedback control mechanism that is implied by Industry 4.0."
Full compliance with standard protocols of communication allows surface vision systems to talk to one another and transfer data. At higher levels, defect images and defect features are aggregated in databases and used in various reporting or Business Intelligence applications.
For example, AMETEK’s designed software sits on its platform as an advanced analytics layer, informing decisions on quality grading, end-use allocation and price adjustment. It also enhances aspects that keep insurers happy, such as compliance and safety.
Adds Zyglis: “The metals industry is one of those leading the charge towards BI, because often they have to show compliance with safety standards. ‘In addition, there are often very defined specifications from customers that would lead to a complaint resolution process if defective material was found. This software acts a layer that prevents such claims.”
Automated control such as that provided in ABB’s ControlMaster range has similarly enhanced the ability of the frozen food sector to achieve legal compliance, enhanced sustainability and better safety standards in the face of lengthening supply chains, explains Tatjana Milenovic, food and beverage group vice president at ABB.
The length of modern food supply chains means that maintaining produce at a constant temperature is vital. Yet the rigorous standards a cold supply chain requires, presents problems because the range of temperatures allowed is narrow.
“So, any deviation and the produce will be deemed unsafe and rejected. As such, the cold system must be able to be monitored and controlled from start to finish,” she explains. “Paper controls and monitoring cannot keep up with the precision needed for cold food supply chains, because they can only register an average temperature. Accurately controlling the temperature inside cold storage requires a smart system, because multiple sensors can record and analyse a constant temperature.”
As ever, technological advances come with a balance of gains and issues. The wealth of data that underpins improved control systems means that “instead of using a spreadsheet with about 50 parameters to support that decision, now you have millions of data points”, says Zyglis.
Yet many industries are losing the experienced operators able to make informed decisions about machine behaviour. The goal with AI is that the systems themselves will diagnose problems lessening the impact of the skills gap but manufacturers may need to address better the data vs security issue, advises Zyglis.
“Our systems are directly connected to the process control network in these plants, so a lot of these facilities don’t necessarily want us having an open conduit to that same system, even if the firewalled internet or cellular network is secure.
“We’re still waiting on industry to decide exactly what level of standards and compliance they expect from security protocols in IoT. In terms of Industry 4.0, the question of ‘what are we trying to do here?’ hasn’t really been answered by the continuous manufacturing market. But the industry and key leaders are beginning to take up the challenge.”