SENSORS IN MANUFACTURING -Developments in Sensor Technologies
Industry 4.0 Magazine talks to Paul Stansfield, UK National Sales Manager at IFM, about how the Industry 4.0 agenda is changing the sensor manufacturer’s business model.
One of the real takeaways that emerged from the Industry 4.0 Conference in Manchester in March 2018 was the notion that capturing sensor data is a good entry point for exploring what Industry 4.0 has to offer.
It’s a view which the IFM Sales Manager shares. “The concepts of Industry 4.0 are not new,” Paul Stansfield tells us. “But there has been a lack of uptake in these concepts until now. One of the things Industry 4.0 has done is made industry think about the concepts of the digital world and how to realise them; to think ‘why do we need that data?’.”
Stansfield warns against the tendency to view Industry 4.0 as an “all or nothing” concept. While it is tempting to approach the topic from an IT perspective and focus on the exciting new technologies of machine learning and artificial intelligence, the reality is that really isn’t affordable for most businesses.
Besides, Stansfield emphasises the machine must be learning from the right data. “It isn’t until you’ve got the ‘grass roots’ piece right that you know you’ve got the right data coming through in the first place,” he points out.
Prove the Concept
Instead, IFM has been advocating a bottom-up approach with its customers keen to explore the Industry 4.0 agenda – adding sensors one machine at a time and proving the concept before rolling it out further to additional machines across the facility.
“For us, it’s all about proving the concept and talking about where the opportunity lies. We can start with one machine – maybe the machine that is performing the worst, do a proof of concept and, when we are ready, we can build up in a modular way. Data capture is going to become almost decentralised. Only the information derived from that data needs to be pushed to the top platform.”
By way of illustration, Stansfield shares the example of a customer who retro-fitted extra sensors across its facility’s central services. The subsequent data analysis threw up some surprising insights into machine performance.
Because the facility had added to the machines consuming those services over time, the services were struggling to keep up with the demand from each machine. Stansfield explains, “We worked with our customer to identify how they could produce in a certain sequence to keep demand lower – so the machine speed wasn’t affected by the constraint those services represented.”
The company now has plans to increase service capacity in order to enhance throughput. This type of data capture and analysis demonstrates how sensor data can offer new – and unexpected – insights.
“It so happens that it is the companies that have some performance issues, where we can really talk about what those issues are and the ways in which we can track performance,” Stansfield says. “Once the concept is proven, it can be escalated and can drive OEE throughout the facility.”
Manufacturing as a Service
Stansfield characterises IFM’s UK market as equally split between end users and machine manufacturers/ OEMs. Given the opportunities for end users to drive improvements using sensor data as a key part of their Industry 4.0 agendas, it might be expected that this ratio might be changing.
However, Stansfield admits the 50/50 split remains steady.
“Many of our OEM customers are grappling with changing business models as well,” he explains. “They are being asked by their customers to provide capital equipment as a service. When you’re charging per piece made or per part delivered, you need to know about the health and conditions of your machines on a continuous basis because if the machine is underperforming, the revenue drops.”
The big piece of this work is monitoring whether the machines are being used within the specified parameters and, in many cases, IFM is exploring with its OEMs how sensor devices can be added to machines to make this data accessible remotely.
Moreover, the data from multiple sites can be used to look at how to improve OEE on a continual basis – which isn’t something OEMs have necessarily prioritised before. As such, demand is growing on both the end user and OEM side of the business.
“Sensors were on machines for control purposes only,” says Stansfield, “however, now our customers are putting sensors in place just for analytical purposes; just to bring back data to meet analytical requirements. It’s changing our business model hugely. Instead of most sensor deployment being for machine control, now we are overlapping technologies and bypassing machine control – going straight to the IT platform and just to gather data.”
Protecting Machine Performance
Stansfield makes the point that, when you ask people what data they are capturing, more often than not the data comes from a PLC controller. Inevitably, that data is very limited. To truly represent the physical world and do true analytics, you need more information than that.
While there is a lot of intelligence in the sensors on machines, often that is not utilised; manufacturers don’t want to take up capacity on their control networks or on their PLC just to analyse data and, consequently, risk network or machine performance.
Cognizant of this brake on deployment, IFM were one of the founder members of the I/O Link consortium. Stansfield believes this new standard will play an important role in the Industry 4.0 journey – enabling organisations to more easily harness the information coming from intelligent sensors.
Stansfield says, “As a founder member of the I/O Link consortium, IFM has played an important role in developing this important standard and we’ve been putting I/O Link technology into our sensors for over five years now. IFM analysis earlier this year estimated that there are 3.5 million IFM sensors already in the market with I/O Link capability. These embedded chips are our preparation for when people want to come onboard with Industry 4.0 – they cover around 80 percent of our product range at this stage.”
About I/O Link
O-Link is the first standardised IO technology worldwide (IEC 61131-9) for the communication with sensors and also actuators. The powerful point-to-point communication is based on the long established 3-wire sensor and actuator connection without additional requirements regarding the cable material.
IFM is hoping this preparation will help organisations that wish to explore what intelligence they can get from sensor data.
“The I/O Master is the only important bit to which people now need to upgrade to gather all the intelligence from the I/O sensors they’ve been buying for the last five years,” Stansfield says. “Probably only around five percent of that data will need to go to the machine controller. 95 percent of it doesn’t need to go to the controller; it will go straight to the IT platform. By doing this, we protect the speed and resilience of the manufacturing process.”
An I/O Link master translates the sensor data to a known industrial standard to send it directly to an IT platform on an industrial network or an IT network for processing and analysis – thereby minimising the extra traffic to the PLC controller and those elements of the network.
In this way, Stansfield argues, I/O Link answers an early objection to getting started with Industry 4.0 and opens up the possibility of collecting the data necessary to truly represent the physical world virtually.
“In this way we can start dividing the physical world into virtual capacities,” Stansfield says, “and these insights will improve our understanding of how well our facilities are performing over time – with sensor data driving OEE. Of course, different companies have different requirements but this is a good, low-cost starting point for everyone.”
Making ‘Cost to Fail’ a Thing of the Past
Richard Jeffers, Technical Director for Northern Europe at RS Components, explains how Industry 4.0 technologies are creating opportunities in new realms of the factory.
Having worked in the automation-intensive brewery sector for fifteen years, Richard Jeffers has been talking about using data to drive decision-making for some time. However, an increasing awareness of Industry 4.0 concepts and associated technologies means he now finds himself pushing on an open door.
“The cost of sensors is coming down and, together with the opportunity that wireless presents to avoid additional cabling costs, this is putting the opportunity to use data to drive decision-making into more hands,” Jeffers tells Industry 4.0 magazine. “That’s really exciting.”
As a leader in the automation space, RS Components is well-placed to have conversations with manufacturers about the opportunities Industry 4.0 presents. Jeffers has had detailed conversations about Industry 4.0 with more than 60 customers over the last 12 months.
These conversations have made it clear that organisations are starting from very different points, Jeffers tell us. Nevertheless, there are some clear stepping stones along the Industry 4.0 journey.
“In a large manufacturing facility there are probably hundreds of sensors and data from those sensors is flowing across the network. The majority of people recognise that there is already a lot of data swilling around their organisation and they want to know how they can use that data.”
Leverage existing data
In these situations, Jeffers recommends organisations ask:
The reducing cost and improving connectivity of sensor technology means it is today far easier to fill those data gaps.
“More often than not, you’re just going to be adding a few sensors to a piece of machinery and that data can go through your existing wired infrastructure to an existing SCADA,” Jeffers says. “If you are going to have to add 100 new sensors to get the information you need, then perhaps you wouldn’t want to put that over the existing network but, again, that depends on how much capacity you have on it.”
Two parts to the puzzle
Jeffers explains that most Industry 4.0 initiatives he has encountered are focused on two areas: either preventative maintenance or process optimisation. “There are undoubtedly other areas,” he admits, but these are the key areas for our customers and, therefore, us. What we’re most interested in is repeatable use cases.”
This means investing time and research into developing models that can facilitate useful equipment and process analyses using cloud-based solutions.
“Ultimately, we envision a model which is ubiquitous for each piece of equipment, whichever hardware you’re using,” Jeffers explains. “Unfortunately, there is an idea that you can just plug a sensor into the cloud and the magic happens. But the reality is far from that. A cloud infrastructure isn’t an end in itself; you need the application knowledge behind it. What are you going to do locally and what are you going to send to the cloud? How are you going to send that data to the cloud? And, once it’s there, what are you going to do with it?”
Clear business case
Ultimately, like any investment, Industry 4.0 initiatives must have a clear objective and targeted ROI.
Jeffers says, “Industry 4.0 is sometimes mistaken for a shortcut to running your factory properly. But it isn’t a shortcut; it’s about using emerging tech to take what you’re already doing to the next level. It is the people who are already competent who are seeing the real value achieved.”
He points to the example of a water treatment works he recently visited, where more than ten years of hard work on OEE has gone into improving operations. Industry 4.0 technologies have provided an opportunity to better understand onsite assets through the development of a digital twin. Subsequent analysis has resulted in a five percent reduction in chemical cost – with a value into the hundreds of thousands of pounds per year – in an already optimised system.
“Digital twins don’t have to be big, complicated projects,” Jeffers says. “Anywhere there is a degree of complexity in a system a digital twin can add value by predicting performance and helping take cost out.”
Jeffers believes that as the cost of solutions come down, digital twins will enable modelling on small-scale operations in the same way that reducing hardware costs are making more analysis possible for condition monitoring to be more widely applied now.
“The lower cost of sensors and cloud computing are making it cost effective to predict failure and optimise performance in areas where it wouldn’t have been cost effective before,” Jeffers enthuses. He points to the example of small electric motors where, until now, it hasn’t been cost effective to add sensors.
“As well as being huge energy consumers, they would run on a cost-to-fail strategy,” he says, “Condition monitoring wasn’t cost effective. As the costs come down, cost to fail will become less attractive. And there are benefits in terms of resource optimisation and environmental benefits to be gained at the same time.”
Ultimately, he says, captured data needs to be leveraged to create insights. And those insights must be actionable.
How do I start with Industry 4.0?
“I’m asked a lot ‘how can we do Industry 4.0?’. There is still a lot of confusion about what it is,” says Richard Jeffers, Technical Director for Northern Europe at RS Components.
He recommends starting with a clear problem that needs solving.
The challenge for solution providers is to create a ubiquitous model for subsequent cloud-based data analysis. Jeffers says this modelling is something with which RS Components is currently grappling: “We’re at an exciting stage; people really want to engage with us.”