Getting the intelligent edge

The hottest thing in computing right now is “the edge” also called “the intelligent edge”. I know many readers will be scratching their heads. The edge? Let me explain.

Think of the edge as the galaxy of sensors being deployed across industry and society. There are sensors in public dustbins to trigger collection when they are full or traffic lights which use sensors to regulate flow. Buildings are studded with sensors to track meeting room usage, light, temperature and all sorts of other things. The Deloitte office in Amsterdam has 30,000 sensors.

The key thing is that vast amounts of sensor data have to be processed locally, near the data source. When there’s no time to send data to the cloud and wait for a response, then the work must be done on site. That’s why it’s known as the intelligent edge.

Autonomous vehicles (AVs) are a terrific example of why this technology is so badly needed. As an AV weaves down a city street it needs to make decisions in milliseconds. Break, swerve, accelerate; it’s a job to keep on the road. If the AV waits for a transmission to arrive from a remote data centre, it’ll wind up bonnet deep in a brick wall. Not even the forthcoming 5G networks, with their ultra-low latency, are going to be enough. The AV needs to have the computing power to process all the data either on board or it has to be nearby at the roadside. It’s the only way. This is possible with the intelligent edge.

There’s also the bandwidth issue. The explosion in sensors is generating huge volumes of data. For example, Walmart is the world’s largest retailer and its entire transaction database contains 2 petabytes of data. One self-driving car will generate that in a single year. That’s just one car. There will be ten million of them by 2020. And there will be many billions of sensor-equipped devices of all kind in a few years. There’s no way existing networks can handle that volume of data.

Computing the data locally is the best way of avoiding clogged networks. Sensor-equipped devices can talk to each other through direct local connections; this is known as mesh computing. It’s a solution to an urgent real-world problem.

So that’s what the intelligent edge is. Here’s why it’s so exciting. Edge computing means we put intelligence where the action is. The effect can be startling.

For instance, HPE equipped the Levi Stadium, home to the San Francisco 49ers football team, with a range of edge technologies. Imagine this. As you drive to the stadium, you are told where the best parking space is; no driving around lost. As you walk to your seat, live data on your smartphone directs you with minimal delays. If you want food, the stadium knows where you are. A hot dog and drink can be brought straight to your seat. When you need a comfort break you’ll know which convenience has the shortest queue.

The potential of the intelligent edge will be determined by our imaginations

Sensors are picking up activity, turning the entire stadium into an intelligent infosphere. This is what happens when the edge is scaled up. And it’s a reality. The concept received a fabulous reception at last year’s Super Bowl. The intelligent edge in action.

This is just the beginning. We are going to see edge devices appear across society. Offices will use sensors and location-based services to optimise desk usage. Just sit down at any terminal and it will be configured for you, ready to use. Need a meeting room? Sensors will detect which room is free. The business model is obvious because desk space can cost £10,000 a year. The return on investment is going to be rapid.

The cost of edge devices is plummeting. I love to collect examples, and I own an internet-connected toothbrush and spoon. They monitor how you use them. Yes, the devices are a bit silly. But they illustrate the economics. Sensors are so cheap they can be installed in a toothbrush or spoon without raising the price.

Naturally companies want to know what’s going to happen with the edge. For example, will it replace data centres? I don’t think so. Instead it’s clear that certain tasks are best suited to specific technologies. The edge is ideal when time and bandwidth are tight. When information is sensitive, then an on-premises data centre is the right place to store and process it. And for big-volume work, the cloud will also remain relevant.

This last point needs stressing. The data generated by edge devices is going to be immense and incredibly valuable. Companies will need to aggregate the data and then interrogate it to find insights. Edge devices can do the preliminary work, generating metadata and so on, but the data centre or the cloud is where the intensive machine-learning and data analytics can be done.

The potential of the intelligent edge will be determined by our imaginations. As a company we’re working on a number use cases, including predictive maintenance, asset management and tracking, intelligent spaces (buildings and venues), healthcare, future cities and connected vehicles. We can create health sensors for the elderly in retirement homes or monitors for heart conditions and other time-sensitive maladies. In India, there is a need for better ante-natal care. Intelligent sensors can help doctors in another part of the country offer a diagnosis.

The intelligent edge will connect industries. The supermarket industry needs to think about how it can make the most of the connected car. How about sending a discount to a car as it arrives? This is the sort of proposition retailers and manufacturers are facing.

This a new era. We are opening the box on a technology and exploring what we can do. The edge can make our world more efficient and effective, and create a more pleasing, greener environment. The word revolution gets used a lot. This time it’s deserved.

To find out more please visit:  https://www.hpe.com/uk/en/solutions/internet-of-things.html