Edge computing has four essential layers: sensors, gateways, edge compute, and cloud, which are unique to each organisation. Different use cases demand different resources at each layer, depending on desired functionality.

IT architectures are therefore expected to be packaged and customised by IT teams, to realise the unique benefits of edge computing in the context of business needs and the layers highlighted above.

The whole point of edge computing is that it is a distributed, open IT architecture, which is designed to decentralise computation and storage.

Thus, it enables mobile and local computing and the Internet of Things (IoT). Data is processed by devices, local computers and local servers, not external data centres or cloud computing platforms. This architecture model is fundamentally different to that which large enterprises have adopted with data-processing warehouses.

Because edge computing enables data-stream acceleration and real-time processing, without eating up bandwidth and network resources, it creates a more agile IT environment by allowing data to be accessed and programs to be run almost instantly. Real-time and context-driven intelligence can empower agile and proactive business decisions, boosting an organisation’s opportunities considerably.

But it is with the processing of large amounts of data where edge computing becomes a particularly useful architecture. Because data is processed near the source, bandwidth is reduced, as are business costs. This helps achieve better return on investment (ROI) and can create better client and customer experiences.

This then leads onto the IoT. Edge computing is perfect for the IoT because firstly, data is processed near the point of origin, or source, which reduces latency between devices and data. Secondly, because edge computing localises computation and data, if one IoT device should happen to malfunction, this won’t have a knock-on effect. Another device will simply take over to maintain availability and reduce downtime.

Whether businesses have an open or closed IT architecture, edge computing should be approached with a view to considering how edge devices will fit into the larger architecture in the context of evolving business needs.

Something to bear in mind though is that because edge technology is advancing so quickly, and there are so many potential use cases, it can be impossible for IT architecture to keep up on a large scale. Instead, some will find running their IT as a service (i.e. running aspects of the things they do today as a service) to be a good model to get into edge computing and start reaping some if its benefits now.