Living – or Driving – at the Edge: Bringing Computing Closer to Home
Among baby boomers, there is a familiar phenomenon called boomerang boomers. They leave home, travel the world, seek their fortune – and sometimes, return home to Mom and Dad, like a boomerang, when things don’t work out too well.
Computing is in some ways like a boomerang boomer. Amazon – and the prescient Jeff Bezos – pioneered cloud computing 12 years ago. Cloud computing delivers computing services over the Internet (“the cloud”). Companies offering these computing services, cloud providers, typically charge for cloud computing services based on usage. The software, applications and data storage may be very far indeed from the client, in huge distant server farms. Clients no longer need to have all their applications, software and data storage on site.
Now, computing is coming back, closer to home. Edge computing takes part of the cloud computing services and brings it home, to the “edge” of the Internet, usually a place where it makes contact with the physical world or end users. It does this by creating edge “cloudlets” (little clouds).
Why do this? Edge computing can save micro-seconds, the time it takes for data to speed to the Timbuktu Cloud and back. This is vital for self-driven cars, where split-second decisions are needed. It can save disastrous failure – imagine, 4,000 self-driven vehicles rely on a distant cloud server, which fails – leaving 4,000 ‘headless horseman’ vehicles that each has lost its brain. It can save cost – transporting data back and forth causes bottlenecks and can be expensive.
Edge computing is often peer-to-peer, person to person, rather than person-to-cloud-provider-to-person. And its arrival could not have come at a better time. According to a report by CB Insights , “the [autonomous vehicle] industry will soon reach a tipping point, where the volume of vehicle data generated will overwhelm existing cloud computing and communications infrastructure resources”. Why? According to Intel, each self-driven car generates 4,000 gigabytes (4 terabytes) daily. Now, Intel has a clear commercial interest in exaggerating this number (it makes ever-more-powerful processors for self-driven cars)… but even half that number is enormous. By definition, self-driven car technology is peer-to-peer. So edge computing is ideally suited for it.
Why the Clouds Are Becoming Darker
There are about 280 million vehicles in the U.S., almost one for every man woman and child. Suppose, let’s say, 10% of them are self-driven, by 2024. That’s an underestimate. 28 million times 2 terabytes of data daily – that is an enormous amount of data to send to the cloud and send back, seamlessly, quickly, without any error tolerated ever. Existing cloud infrastructure is simply not adequate and new a combination of edge and central cloud computing will need to co-exist in a connected and autonomous future.
Posted by Noam Maital, CEO at Waycare
 CBInsights. “What is edge computing?” Available online at the CBInsights website.