Data is the true currency of Industrial IoT tech

The benefits to supply chains of Internet of Things (IoT) technology are substantial, and include increased visibility, more collaboration, better customer service and real-time tracking of goods.

Forward-looking organisations are increasingly turning to IoT networks to transform complex supply chains into fully connected networks. Sensor data and Radio Frequency Identification (RFID) from such devices can deliver real-time asset tracking, monitoring, and alerts that help streamline tasks and minimise disruption. 

A subset of the IoT – which consumer-facing uses such as wearable devices and smart home technology – is the Industrial IoT (IIoT). 

This consists of internet-connected machinery and advanced analytics platforms that process the data generated by IoT devices, which range from tiny environmental sensors to complex industrial robots. 

Although the world ‘industrial conjures images of warehouses, shipyards and manufacturing production lines, IIoT technology is revolutionising a range of industries, including agriculture, healthcare, financial services retail and even advertising.

For businesses that produce or transport physical goods, IIoT tech can offer significant efficiencies, and can help businesses reimagine operational modus operandi. 

In supply chain, for example, sensor-managed inventory can automate the supply ordering process, just before items go out of stock. This reduces waste and frees up employees to focus on more-strategic tasks.

In manufacturing IIoT-enabled machines can self-monitor and predict potential problems, meaning less downtime and greater efficiency.

But IIoT having the potential to help, and it actually helping, comes down to one critical factor: data. Or, more specifically, how data is harvested, and then used.

Andy Hancock is Global VP Centre of Excellence, SAP Digital Supply Chain. The Centre of Excellence is like a global SWAT team of industry experts, serving SAP colleagues in the field, with knowledge transfer. 

Hancock points out that IIoT networks generate mountains of data and warns that powerful 5G technology – with its huge data capabilities – can “make people lazy, and so they end up throwing tons of information around just because they can”. 

He adds: “The trouble is that when you scale-up this approach to enterprise level you soon end up with 50 million data points that flood the network, making it inefficient. 

He adds that the danger is that businesses who fall into this trap then end up “chucking more technology at the problem”, where really what they should be doing is “coming back to fundamentals”. 

Hancock says the secret to working with big data, like that generated by IIot networks, is “to always be looking out for exceptions”. 

He adds: “Think of a temperature gauge on a piece of equipment that is feeding back data. As long as everything is running okay, the equipment will always be roughly the same temperature.

He says: “You don’t need to keep feeding back data about that piece of equipment. The only data you want to capture is if something changes – for example, if the thermostat fails. 

“IIoT devices are creating data 24/7, so the idea is you discard most of it and look for the exception - the piece of data that shows a machine is overheating, or out of calibration. With data, less is usually more.”

Hancock says businesses who are harnessing IIoT tech should treat their networks in the same way the world treated modems, back when the Internet was in its infancy.

“Think back to the days of dial-up modems,” he says, “where everyone minimised the amount of data transmitted, because if you didn’t then the whole thing just hung”. 

As well as focusing in on data exceptions, Hancock says IIoT operations benefit hugely from edge computing, which is a distributed computing model that brings computation closer to the sources of data – a warehouse full of inventory, say, or components in a vitally important production plant.  

“If you have local processing power it can be invaluable,” Hancock says. “For instance, if a machine is going out of tolerance at a paper mill and there was a delay in sending this data to the cloud and back, then you could have lost a hundred metres of product by the time the machine is switched off. 

“So the idea is that local machine-learning tech understands that the machine is out of tolerance, and recalibrates it without any human action needed.”