Edge computing is far from a new concept, but with the arrival of 5G communications technology, it is able to fulfil its potential. By leveraging the ubiquitous bandwidth of 5G networks, data collected from devices at the network edge can be made available to every enterprise system, giving businesses a complete, near real-time overview of their entire value chains.
What is edge computing?
Edge computing is essentially about putting compute capabilities close to where data is generated – on oilrigs in the middle of the North Sea, in the base of wind turbines on rocky outcrops of land in remote areas, or inside machines on a factory production line. In many cases, the sheer volume of raw information collected has meant that sending it all to the cloud for analysis has been prohibitively expensive (consider that an average connected car sends 25 GB of data every hour). Also bear in mind that in many cases, the network edge is far from heavily-populated areas and therefore also not so likely to be served by a mobile phone and data network.
This has necessitated the strict filtering of data sent up to the cloud for processing – which in turn has restricted the level of compute power available to perform predictive analytics on data coming from IoT devices, such as monitoring the performance of assembly line machines or components, to ensure they receive maintenance as and when needed, therefore helping avoid unplanned downtime.
Data processing at the network edge is back
This has led to the resurgence of local compute power – something that mainly died out over the last 10 years, as enterprises consolidated their processing into central data centers – and moved from on-premises to cloud computing. Now, data processing at the network edge is back – often in the form of ruggedized servers, like Fujitsu INTELLIEDGE. These servers use low-level communications such as Wi-Fi or Bluetooth to collect data from nearby devices, sensors or other connected ‘things’, and can undertake very low-latency data processing. This allows decisions to be made very quickly – without the cumbersome process of needing to filter data, send it information to the cloud and then wait for results to come back.
However, we are now beginning to understand that there’s more potential value to this data, beyond the edge. Through connecting this information to enterprise networks, we can build a clear picture of entire value chains.
How 5G is unlocking data potential
5G is the technology that will help unlock this potential by bringing vast bandwidth that will allow edge servers to share all the data collected from the edge.
Bringing this data into the center of a network means it is available to the heavyweight number crunching capabilities and advanced analytics of neural networks. What’s more, when it is combined with existing information such as customer records, sales numbers or even the weather report, neural networks can uncover previously hidden patterns or correlations. As a result, businesses achieve visibility over entire value chains, can respond more rapidly to market changes and are better placed to identify new opportunities.
The emergence of 5G also means that business operations relying on regular communication between edge devices and central corporate servers can become more resilient. Let’s take one of our European customers for example: A restaurant chain we provide with technology and management services. In each branch, staff use tablets to take orders. These communicate with electronic point of sales devices (EPOS) via Bluetooth, and the EPOS talks to a centrally-held database via a wired LAN connection.
When any of these connections fail, it causes problems. 5G networks offer an alternative network route at all times, meaning that in case of a hardwired network outage, the tablets used to take orders – which are right at the very edge of the network – can shift to communicating via 5G mobile networks. This provides resilience. For customers without existing fixed connections, 5G could deliver the connectivity needed for their primary network, and avoid all the cabling in the process.
The huge bandwidth offered by 5G combined with the ability to add sensors to almost anything means there is huge potential for leveraging data. From devices that track our vital signs and send these direct to doctors for assessment, to fully autonomous cars that are able to react appropriately to constantly changing conditions, the army of connected ‘things’ is growing exponentially. In fact, Ericsson predicts that there will be 29 Billion connected things by 2022 – of which around 18 billion will be IoT related. The company also anticipates that soon after, in 2024, up to 65 percent of the global population could be covered by 5G technology. Consequently, the possibilities for collecting data, and deriving new value from it, really are endless.