Industrial Technology - Linked-in Industrial Technology - Twitter Industrial Technology - News Feed
Latest Issue

Big Data versus ‘smart data’

With the Industrial Internet of Things enabling data to be collected from ever greater numbers of devices, the question of just what we do with all of this data has become a real connundrum. Turck product manager Sander Makkinga explains the concept of smart data.

While the terms ‘the Industrial Internet of Things’ and ‘Industry 4.0’ are related, they are not the same thing. One can certainly be an enabler for the other, providing the mechanism by which users collect the data that is needed in order to make better production decisions. And certainly Industry 4.0 strategies rely on information. But is there a risk that we could get so bogged down in data that we lose sight of what it means?

“Consider, for example, vibration monitoring on a machine,” says Sander Makkinga, product manager at Turck’s Mülheim an der Ruhr headquarters in Germany. “That could be generating terabytes of data, and perhaps within that you’ll notice a spike. Should you worry about that spike? Has it affected production? Should you stop the machine? Is it part of a trend or was the sensor just picking up somebody closing a door?”

The point is that you could spend hours poring over these terabytes of data without there being any evidence of a production issue. This is Big Data, and there’s a role for that. But from a production perspective, is it ‘smart data’?

Industry 4.0 sets out a strategy for flexible production processes that create value for all stakeholders right through the supply chain, not just for the manufacturer. But in particular it delivers highly integrated, highly flexible production capabilities for businesses to meet the individual and customised requirements of customers, right down to ‘batch size 1’. “The enabler for Industry 4.0 is the merging of industrial processes with IT processes, allowing companies to optimise production by giving them complete plant visibility,” says Makkinga. 

Industry 4.0 is built on data, from which manufacturers can optimise processes to boost quality, productivity, efficiency and availability. Appropriate data tied in with control systems and higher level business systems gives manufacturers a new level of agility, enabling them to achieve faster delivery times on smaller batches of products customised to a wider range of individual customer requirements.

With modern networking technologies such as 

IO-Link, Industrial Ethernet and OPC UA, data can be transferred even from the humblest field sensor up to ERP level and beyond to the cloud, giving manufacturers all the information they need. This forms the basis not just of process optimisation, but also of predictive maintenance. But extracting the smart data from the Big Data is what will really drive the marginal gains that will impact most on the bottom line.

“As an example of the power of smart data, consider an air-actuated valve,” says Makkinga. “That valve has two positions, open and closed. Before we started discussions about Industry 4.0, all the user might have monitored was the open/closed signal. But suppose you connect that valve to a fieldbus module that has an internal clock. Now, not only can you see whether the valve is open or closed, but how long it took to reach either position. 

“Let’s assume that when the valve was new, it took one second to move from its closed position to its open position. Two years on, you notice that it’s taking longer, increasing to 1.5 seconds, then 2 seconds, then 2.5 seconds. That’s indicative of a problem, and already you can start to think about maintenance. You can base a decision on how quickly the open/close time is lengthening: if it’s happened over the course of a couple of days, you might want to schedule in maintenance to swap out the valve quite quickly. If it’s happened over the course of a month or more, you might just want to monitor it for a bit longer.

“But we can go further. Perhaps the valve is one of ten  similar valves. Is the problem limited to a single valve, or is it affecting all of them? If it’s the latter, then the problem might not be the valve at all, but rather with the compressed air supply pressure. The point is that, with no more effort than choosing exactly what data to monitor and then interpreting that data, we can make much smarter decisions about maintenance, and potentially significantly boost the reliability and availability of our systems. No additional sensors have been used – just a combination of an open/closed signal with a time stamp. It’s just how you use the information and how you combine it with your application knowledge.”

This is making use of smart data for real production improvements, rather than collecting big data for big data’s sake. “Businesses need to be able to decide what level of information is useful for their applications rather than just collecting everything,” says Makkinga.

Turck provides many of the enabling technologies to deliver on this new paradigm of smart data, including sensors, PLCs, HMIs, I/O blocks, field logic controllers, monitors and more. An established player with Industrial Ethernet and IO-Link communications, the company is also at the leading edge of OPC UA, as well as offering innovative multi-protocol technologies.

Download pdf

Other News from Turck Banner

Moving from centralised to decentralised automation

Latest news about Machine Building

Additional Information
Text styles