The Evolution from IoT to Edge: A Conversation with Shekar Ayyar
We’ve all heard it said that the pace of tech innovation gets exponentially faster every day. And one could say that edge computing is the posterchild.
First named in 19991, the internet of things (IoT) meant any device connected to the internet. Early use cases included simply checking if a device was on or off. Yet IoT was swept along with advancements in other technological systems, like cloud, machine learning, cloud-native apps and mobility. And as use cases grew in complexity, like smart cities and homes, and the number of devices proliferated, IoT became a component within a bigger picture: edge.
- The evolution from IoT to edge.
- How the pandemic and other 2020 events impact adoption and innovation.
- Edge use cases for today and tomorrow.
- The infrastructure and strategies that organizations need to make edge come to life.
A few years ago, the big buzzword was IoT. Now, it’s “edge.” How has this concept evolved over the past few years—and why?
When we started talking about IoT, it was mostly about devices and endpoint diversity. What’s happened since is that most people have recognized that the cloud world, which is the infrastructure behind this, is also changing. And it’s not just the device that was sitting at the end. It’s also the infrastructure connecting those devices to the backend infrastructure of compute, network and storage. Specifically, this is now much more distributed than originally conceived.
If you wind the clock back, we started with a private cloud infrastructure. Then, the pendulum swung to public clouds and how they’re serving customers. And now, we’re swinging the pendulum back, saying we do need compute power closer to the point of consumption. And we need these resources at many different points, not just in 25 or 30 consolidated data centers.
So, this change is now what becomes the basis of edge computing and edge infrastructure. When combined with the diversity of devices, this combination is a lot more powerful in terms of how applications are created and consumed by both consumers and enterprise customers.
How has the pandemic and the global shift to remote accelerated interest in and adoption of edge?
Obviously, the pandemic has been a really unfortunate thing that all of us have had to deal with. But alongside that has come this notion of “remote everything.” Everybody’s operating remotely, from how they’re getting work done to interactions between customers and businesses. As a result, there’s a much greater requirement for the network to support connectivity and scale.
On the device side, people want to have immediate access to information and applications. The network is getting stretched to support things like video conferences and other apps being accessed at the same time by many people. The backend infrastructure must support that volume of activity and the bandwidth required.
Finally, there’s this characteristic around latency, which is this notion of asking for something and expecting it right now. At the end of the day, this will play into things like autonomous cars. But right now, with the situation caused by the pandemic, there are several things that we want, simple things like no time lags between audio and video in virtual conferences. Low-latency, high-performance, high-bandwidth networks are going to be required to make this work. All of this is becoming a basis for the future of edge computing architecture.
VMworld 2020: 5G Is Here
Watch Shekar Ayyar's Solutions Keynote at VMworld 2020 on-demand.
How are organizations using edge now? And what are some of the innovative use cases for the future?
With the combination of cloud computing and communications infrastructure, further catalyzed by 5G, this is now becoming a high-performance infrastructure. There’s a subset of things that people are talking about today regarding applications like video conferencing and touchless ATMs. But that then leads us to think that this is going to touch every walk of life that we’re all experiencing.
Like the internet is now becoming the necessity that we cannot live without, we’re going to have everything from retail to manufacturing to financial services to transportation to healthcare—all of which will have rich use cases delivered on top of this infrastructure. Examples are things like remote medicine.
Today, people think about placing a call to your doctor, so you don’t have to go into physically meet them. But there are many, many more sophisticated applications, like remotely delivered telemedicine in the form of surgery. Robotic arms are being maneuvered on mobile devices.
In industrial automation, you can remotely manage an entire factory from a phone or an offsite control center. For example, you could have an automobile unit built remotely so an individual doesn’t have to go visit the factory.
So, the raw materials for building these applications are now becoming real and available. The whole goal is that you can take things like geolocation, network slicing and qualities of service and make these available as API-driven architectures that developers can build applications on top of. It’s very similar to how the cloud operates.
Furthermore, organizations will be able to almost dream up capacity wherever it’s needed and then acquire it at that point. In the future, you’ll pay for the right-sized amount for whatever quality of service you want. If you want low-latency, high-performance, pay a little more and then you can get that quality of service for your users.
Why do customers look to partners, like VMware and Dell Technologies, to help bring their edge vision to life?
The combination of VMware and Dell Technologies is an excellent complement for getting industry-leading hardware platforms and software abstractions. Together, we help our customers extend from the central data center to regional data centers and all the way to endpoints. As an example:
- We bring racks of servers loaded with software to operate an entire data center in an Infrastructure-as-a-Service format.
- We then have Tanzu from VMware layered on top of it, which becomes the application layer.
- Then, we can look at Telco Cloud to take the network of these telecom operators and have all of their network and containerized functions operate on top of that.
- We can then take this cloud infrastructure and extend into the edge with smaller versions of these storage, compute and network data centers. These can operate closer to where people want to consume capacity to enable low-latency, higher bandwidth and proximity computing.
- Further, we can then take that to the network edge with VMware VeloCloud SD-WAN combined with Dell hardware. This allows for the entry of edge on-premises, which extends to the actual endpoints with Dell clients and VMware Workspace ONE
- Finally, we layer on intrinsic security and the ability to manage everything remotely.
Combined, this end-to-end architecture enables customers with an edge infrastructure that no other two vendors can together provide.
1Foote, Keith D. “A Brief History of the Internet of Things.” Dataversity, August 2016: https://www.dataversity.net/brief-history-internet-things/