What you need to know about fog computing
Computing infrastructure is beginning to sound like a highly changeable weather report. Sunny-day networking scenarios involve good cloud cover...
Just as businesses get a good handle on the cloud, we welcome edge computing: the next must-have in data processing power.
Edge computing is being touted as the next multibillion-dollar tech innovation that will take the place of cloud computing and help power the best benefits of the Internet of Things (IoT). But for many experts, the edge will be the sharp line of the cloud’s processing power, taking some data storage and analytics on at the edge of the network, keeping the cloud free for deeper analysis.
Edge, or fog computing as some manufacturers prefer to call it, is the answer to a simple problem. As IoT devices multiply, the volume of data to be analysed will increase exponentially. At the same time, to fully realise the benefits of IoT sensors – getting real-time analytics and solutions on real-time data – processing must be lightning fast. Waiting for the data to be transferred to the cloud, processed, analysed and an answer returned will just take too long.
Let’s take a real-world example. Although many companies are currently investing in self-driving vehicles and the first of these cars are on the road, cities full of self-driving cars are a futuristic dream until data from sensors can be processed in real-time. As a self-driving car approaches a traffic light, it should be able to read data like location and speed of other approaching vehicles, cyclists and even pedestrians. It should also know the exact moment the light will switch from green to red and calculate the correct speed at which to approach it. But all of that data has to be scooped up and analysed in an instant. That’s where edge computing comes in.
Edge computing provides a layer between the network and the cloud. By putting processing power into local computing devices, specific data can be analysed right then and there, instead of being sent to the cloud.
“The agility of cloud computing is great – but it simply isn’t enough,” says Thomas Bittman, vice president and distinguished analyst with Gartner Research. “Massive centralisation, economies of scale, self-service and full automation get us most of the way there – but it doesn’t overcome physics – the weight of data, the speed of light. As people need to interact with their digitally-assisted realities in real-time, waiting on a data centre miles away isn’t going to work.”
There are a number of physical technologies currently being explored for edge computing, including gateway networking devices, industrial PCs and micro data centres. Mark Darbyshire, VP of platform and integration at SAP, tells Think Progress that APIs will be a key driver of the edge.
“API management has been proven to work over the last few years,” he says. “Everyone uses APIs to expose their capabilities to the outside world because it allows them to do it in a way that’s secure, manageable and they know who’s asking them for what. They can also decide whether they want to do it or not, and it’s easy and people will leverage that existing architecture. But I think increasingly people will take their APIs and micro-services approach and say: ‘Why can’t we have micro-APIs and micro-gateways as well?’”
However it’s implemented, enterprises have already started adopting edge computing. Research firm IDC predicts that by 2019, at least 40 per cent of IoT-created data will be stored, processed, analysed and acted upon close to, or at the edge of, the network.
“People are starting to embrace it,” Darbyshire says. “I think when you start to do major IoT projects, you realise this isn’t about bolting something onto your business. This is about connecting two previously separate parts of your business.
“You don’t connect things by just throwing them together. You have to keep them distinct. You connect them through a line of interaction, and that’s the edge. Today the edge is quite short, but over time I suspect the edge will extend and become more profound.”
Gartner’s Bittman agrees: “The time to have an edge strategy is very, very soon. Watch as VR goggles start to take off, and heads-up displays in cars, and mixed reality apps on smartphones (doing more than Pokémon), and maybe Google Glass II. These technologies will push the edge explosion, soon.
“The cloud will have its role, but the edge is coming, and it’s going to be big.”
For a long time, enterprises have been focused on the cloud – debating about public, private or hybrid architectures, pondering the challenges of data privacy, security and regulations, as well as latency, bandwidth and downtime. Edge computing offers potential solutions to many of these challenges and an opportunity to take full advantage of an IoT driven by mixed realities, deep machine learning and powerful data analytics.