The recent COVID-19 global event and the ongoing march toward concepts such as Industry 4.0 are showing us just how important connectivity is to our digital lifestyle. During the pandemic, video streaming, gaming and conferencing demands exploded as we worked and studied from home. We watched more Netflix, conferenced with our families or work colleagues and gamed with the global community. Networks didn’t flinch. After all, the internet has always been focused on delivering information to humans, and we will always hit limits on the amount of information we can consume. We humans can only process so much information, and there are only so many of us on the planet. But what happens when a consumer internet comes up against the internet of things (IoT) and Industry 4.0? Machines don’t have eyes – and they certainly don’t take holidays – so massive machine-type communications (mMTC) will continue be an increasing driver for networks.

Not all mMTC is created equal

There is virtually no end to the amount of information machines can consume. And unlike our network usage, their communications aren’t mostly unidirectional – the amount of information they generate is equal to or greater than what they receive. Artificial intelligence (AI) and machine learning applications create streams of information from a huge variety of connected devices and sensors. It’s a lot for networks to process.

But not all mMTC applications need the same type of connection. A connected vending machine or a real time control application doesn’t need the same kind of connection as a cell phone call.

So, in a word unpredictability. There is no limit to the number of devices and applications that can be deployed, and there is no limit on the amount of computing power that can be utilized – and distributed – to process the information. Access networks will become increasingly critical conduits to collect and aggregate all that information into the cloud to be processed (the learning part of ML). That information then makes AI systems smarter and capable of better decisions, which must then be distributed to countless devices that can make instantaneous local decisions (the “inference” part of ML). An example of such a device is the full self driving (FSD) neural processing chips in a Tesla EV.

Another example of expansion at the network edge was highlighted during the COVID-19 pandemic when we saw a huge rise in online shopping. That trend drove the need for fleet management and tracking, which requires a range of devices and applications to be deployed to efficiently manage a dynamic supply chain.

 We will need to tailor network connectivity to reflect the requirements of the device or application. That’s why capabilities like network slicing will be important – to help networks adjust different slices to the latency, capacity and availability needs of different applications.

Every sector from utilities to education to manufacturing to government is looking to leverage technologies such as 5G as part of their strategies. Today, these industries are using LTE to start their digitalization projects – but this is a start of a journey. The demands of more and more devices supporting more and more applications mean that 5G can’t get here soon enough. In fact, we are already seeing traction for 5G in autonomous vehicles and haulage systems, electrical grids, transport networks, shipping ports, containers, and airports.  

When we think of networks, we think of how we use it. But as machine after machine comes online, they will become the primary users. There’s no need for concern though: with technologies that let them adapt to the unpredictable, networks are sleeping with one eye open.  
 
To read more about the Nokia IP and optical networks that lie at the heart of mMTC, visit https://www.nokia.com/networks/networks-keep-us-going/