Gartner defines edge computing as “solutions that facilitate data processing at or near the source of data generation,” a nice, succinct explanation of this booming technology. If you don’t need it already, you most likely will soon as it’s a key enabler of cloud-based applications, including Internet of Things (IoT) applications and others supporting the digital transformation of business.
“Organisations that have embarked on a digital business journey have realised that a more decentralized approach is required to address digital business infrastructure requirements,” says Santhosh Rao, principal research analyst at Gartner. “As the volume and velocity of data increases, so too does the inefficiency of streaming all this information to a cloud or data centre for processing.”
There’s so much buzz around the edge that it has even spawned its own glossary of terms, under stewardship of the Linux Foundation.
The Evolution of Edge Computing
The movement to edge computing follows the cyclical nature of IT trends. We began with a centralised, mainframe-centric model, then moved to a decentralised model with client-server networks, with distributed computing power. The cloud is another example of a centralised model, but this time it’s augmented by the edge, thus creating a hybrid centralised/decentralised model.
This hybrid model combines the best of both worlds. The cloud can be used for data that requires massive amounts of processing or that does not require immediate attention, for example. The edge will support applications that demand lots of bandwidth, rapid response times and low latency. Examples include real-time decision-making and gathering data from intelligent devices, such as in a healthcare setting. The edge is also useful in meeting compliance requirements around where data is physically located.
How Edge Computing Addresses Performance and Regulatory Issues
While they can take many forms, edge data centres generally fall into one of three categories:
- Local devices that serve a specific purpose, such as an appliance that runs a building’s security system or a cloud storage gateway that integrates an online storage service with premises-based systems, facilitating data transfers between them.
- Small, localized data centers (1 to 10 racks) that offer significant processing and storage capabilities. Ideally, these “micro data centers” are delivered in self-contained enclosures that contain all required physical infrastructure, including power, cooling and security.
- Regional data centers with 10 racks or more that serve relatively large local user populations.
As this wide range of options indicates, it’s not the size of the data centre that defines it as edge, but its proximity to the source of data that needs processing, or those consuming it. With edge data centers nearby, bandwidth becomes less of an issue because data often travels over a private, high bandwidth local-area network, where links of 10G or more are common. The close distance likewise solves the latency issue and organisations can place them wherever they need to for regulatory compliance.
The Opportunity at the Edge
With a sound edge computing strategy in place, organisations will be positioned to take advantage of IoT applications – including those incorporating artificial intelligence and augmented reality – to provide significant business benefits.
For example, let’s look at the applications that can help companies improve customer experiences. Retailers are using the technology to enable digital signage that helps customers find their way and alert them to sales, as well as smart mirrors that help customers virtually try on clothes. Industrial field service personnel are using augmented reality applications that help guide them through complex repairs. Healthcare providers use IoT technologies to power digital health records and telemedicine.
Improving operational efficiency is another driver. Artificial intelligence enables predictive maintenance applications, which drives down maintenance costs in areas ranging from manufacturing to data centres, while reducing the risk of failures. Retailers use RFID applications to help manage inventory and reduce losses. Cities can use IoT applications to monitor busy intersections and control signal lights to help with traffic flow.
IoT applications are also driving new revenue streams, and even entirely new businesses. Uber and Lyft depend on it to match drivers with customers. Logistics companies have launched new lines of business around providing real-time status on cargo, including climate controls. Healthcare providers are offering remote device monitoring and analysis services.
The possibilities for IoT applications are virtually endless, but many if not most of them rely on edge computing to deliver the performance they need.
Protecting Edge Computing Environments
Given their critical role in IoT and other critical business applications, edge data centres must be protected in much the same fashion as traditional data centres. That means providing tools that enable remote management as well as physical security.
In terms of physical security, you need to make sure no unauthorised users can access the compute infrastructure – the first step in providing proper cyber security. A regional edge data centre will likely have security such as card readers on the door. But physical security also needs to be addressed at un-manned edge data centres in remote locations. They will likely require sensors and security cameras that can be monitored remotely and issue alerts for everything from excessive temperature levels to water leaks and unauthorised human access.
Similarly, edge data centres with little to no IT staff on site require remote management capabilities, enabling a centralised group to perform day-to-day management of the infrastructure.
And it’s not just compute infrastructure; you also need to manage and protect the network components that connect the edge data center to the cloud, enabling a hybrid data centre-cloud environment. Without network connectivity, the edge environment is just an island.
Access Edge Computing Resources
Many companies, such as retailers with many stores, also need a strategy for how to deliver a series of edge data centers that meet all of these requirements quickly and reliably.
At APC by Schneider Electric, we have invested significant resources in delivering everything customers need to quickly roll out reliable, high-performance, secure edge data centers, including micro data centers, to allow you to achieve Certainty in a Connected World. To learn more, visit our edge computing page.