AT&T* is reinventing the cloud to boost the potential of self-driving cars, augmented and virtual reality, robotic manufacturing, and more. We’re embracing a model called edge computing (EC) to move the data crunching from the device to the cloud. Driving it will be single-digit millisecond latency that only tomorrow’s 5G can deliver. And powering it all will be our software-defined network, the most advanced of its kind in the networking industry.

Here’s the challenge: Next-gen applications like autonomous cars and augmented reality/virtual reality (AR/VR) will demand massive amounts of near-real time computation.

For example, according to some third-party estimates, self-driving cars will generate as much as 3.6 terabytes of data per hour from the clusters of cameras and other sensors. Some functions like braking, turning and acceleration will likely always be managed by the computer systems in the cars themselves.

But what if we could offload some of the secondary systems to the cloud? These include things like updating and accessing detailed maps these cars will use to navigate.

Or consider AR/VR. The industry is moving to a model where those applications will come through your smartphone. But creating entirely virtual worlds or overlaying digital images and graphics on top of the real world in a convincing way also requires a lot of processing power. Even when phones can deliver that horsepower, the tradeoff is extremely short battery life.

Edge computing addresses those obstacles by moving the computation into the cloud in a way that feels seamless. It’s like having a wireless supercomputer follow you wherever you go.

“Edge computing fulfills the promise of the cloud to transcend the physical constraints of our mobile devices,” said Andre Fuetsch, president of AT&T Labs and chief technology officer. “The capabilities of tomorrow’s 5G are the missing link that will make edge computing possible. And few companies have the sheer number of physical locations that AT&T has that are needed to solve the latency dilemma.”

The faster speeds and lower latency expected with 5G will be key elements to enabling edge computing. But latency is also determined by the physical distance between a mobile device and its network resources.

Say you want to run a VR experience in the cloud. And the data center powering that experience is hundreds of miles away from you and your VR glasses. Every time you turn your head, there’s a good chance there will be a noticeable delay between when you turn and the image follows. That lag is unavoidable because of the time it takes the data to cross large physical distances.

So we’re shrinking the distance. Instead of sending commands hundreds of miles to a handful of data centers scattered around the country, we’ll send them to the tens of thousands of central offices, macro towers, and small cells usually never farther than a few miles from our customers.

If the data centers are the “core” of the cloud, these towers, central offices, and small cells are at the “edge” of the cloud.  Intelligence is no longer confined to the core. The cloud comes to you.

We’ll outfit those facilities with high-end graphics processing chips and other general purpose computers. We’ll coordinate and manage those systems with our virtualized and software-defined network.

We could someday embed these systems in everyday items like traffic lights and other infrastructure. That could enable self-driving cars to talk to their surroundings or alert fire and medical services almost instantly when there’s a problem. You could get amazing VR and AR images delivered instantly to the super-slim device in your pocket. Doctors could view and share and adjust complex medical images without investing in expensive imaging systems.

Edge computing could also spark the next generation of robotic manufacturing. The 5G service on the horizon could play a vital role in what’s called “Industry 4.0 – Digital Manufacturing”. The anticipated low-latency wireless connections could eliminate the traditional wired connections to robotic assemblers. Updates come quicker. Products can get to market faster.

We’re already deploying EC-capable services to our enterprise customers today through our AT&T FlexWareSM service. Customers can currently manage powerful network services through a standard tablet device. We expect to see more applications for EC in areas like public safety that will be enabled by the FirstNet wireless broadband network.

We’re committed to deploying mobile 5G as soon as possible and we’re committed to edge computing. As we roll EC out over the next few years, dense urban areas will be our first targets, and we’ll expand from those over time.

Our network virtualization initiative will go hand in hand with our mobile edge computing program. Our goal is to virtualize 75% of our network functions by 2020. We aim to cross the halfway mark this year, reaching 55%. As we’ve said before, we think 5G and software defined networking will be deeply intertwined technologies. We don’t think you can claim to be preparing for 5G and EC if you’re not investing in SDN.

We’re all in. Now.

Our AT&T Labs and AT&T Foundry innovation centers are at the heart of designing and testing edge computing. In February, the AT&T Foundry in Palo Alto, California, released a white paper on the computing and networking challenges around AR/VR. We’ll put out a second white paper in the coming weeks. It will discuss how we can apply edge computing to enable mobile augmented and virtual reality technology in the ecosystem.

There’s no time to lose. We think edge computing will drive a wave of innovation unlike anything seen since the dawn of the internet itself. Stay tuned.