AT&T and IBM have teamed up to offer hybrid cloud services at the mobile network edge for enterprises in order to give them right-sized solutions. AT&T has turned its network core into a software virtual network function that can work with IBM's Satellite Cloud platform to deliver services and applications to enterprises through AT&T's multi-access edge network (MEC.)
IBM and AT&T are partnering to serve mutual customers, with AT&T being one of the first to tap into IBM's Satellite Cloud. Satellite Cloud is currently in beta proof of concept trials with general availability slated for the middle of next year.
Using LTE, or 5G, enterprise customers are able to design and manage services across multiple clouds to billions of edge devices with high levels of security. IBM brings its encryption to the table, as well as its Watson artificial intelligence capabilities.
"Probably manufacturing is the simplest use case or vertical to think about," said AT&T's Robert Boyanovsky, vice president of mobility and IoT. "So customers in that manufacturing space can use cellular and keep that data local and private.
"When you think about the CSPs running applications, what are they going to be doing on it? The first, and the simplest, is data analytics; AI and ML on the data itself. So just think of lots and lots of endpoints, IoT-type endpoints, getting locally steered into the data center and then having the ability to run their applications locally and drive some insights off of the data. That's probably the first easy thing to do here."
IBM Satellite Cloud provides a single pane of glass to manage and catalog the services from the cloud. The management capabilities are enabled via Red Hat's OpenShift platform, which AT&T used prior to IBM buying Red Hat last year.
"We see this huge opportunity for hybrid cloud and 5G. We think that's a tremendous opportunity for service providers like AT&T to establish significant new value in the market," Steve Canepa, global industry managing director, telecommunications, media and entertainment, at IBM. "We think because of the great pace of change, doing it in an open architecture approach is going to provide tremendous advantages. So what the announcement is really about is our Cloud Satellite capability, which is built on OpenShift from Red Hat."
Red Hat's Kubernetes-based OpenShift allows service providers to tap into automation across on premise, data center and third-party cloud environments. Red Hat orchestrates the software and applications on top of the infrastructure.
"It's (Cloud Satellite) very efficient and can respond in a dynamic way to these changes in consumption patterns that we're seeing, and the changes in the kind of workflows that are happening," Canepa said.
With the combined platform, Boyanovsky said enterprise customers become the owners of their own localized data. Based on the size, resources and skill sets a company has, the tech stack could be managed by the enterprise or by IBM
"Private cellular networks are essentially turning what used to be the carriers' data into the enterprise's data and letting them have their own sandbox to drive their own outcomes," Boyanovsky said. "So it can literally be any business model on that platform. It gets simpler to manage, and because it's localized, it heightens the security posture.
"You have a secure end-point device that hits the in-building cellular system, which could be CBRS or licensed spectrum, it doesn't matter. All of that traffic gets sent to the localized data centers where they compute it. It doesn't go back to the core network."
AT&T and IBM provided a glimpse of what the combined offering is capable of in August. The trial collaboration took place at IBM’s Watson research lab in Yorktown Heights, N.Y. and included AT&T 5G+ (mmWave), IBM's cloud and AT&T's MEC technologies. MEC is a private cellular, low latency solution that can process data on a business premise instead of routing traffic over public networks.
One of the scenarios the two companies were exploring at Yorktown Heights included enabling a researcher to remotely adjust locations of IoT network devices in a laboratory. Another included enabling a systems administrator to remotely rewire machines in a data center to provide a more agile environment.
"Video is probably the next use case that makes sense, via security surveillance," Boyanovsky said "So that could be quality control and safety inspection of parts in real time. That is certainly a viable next set of use cases in manufacturing that we've seen.
"Real time sensors in your cameras provide real time inspection of manufacturing defects, or object recognition and contact tracing. Think about where we are with Covid and some of the use cases that are out there to make sure everyone's staying safe. With a camera and a local tech stack, that can be accomplished."
Boyanovsky said it was still early days for the combined platform, and while AT&T may have an edge now, other service providers will no doubt come up with similar offerings that could also include IBM.
While the IBM press release touted 5G and hybrid cloud, Boyanovsky said the platform was "born in LTE" but will be carried forward into 5G once the new radios are available. While carriers are touting their nationwide 5G deployments, the platform from AT&T and IBM provides a means for monetization now.
"It's great for customers because they don't have to wait for an indoor certified 5G network," Boyanovsky said. "They can start with LTE, whether that's a DAS, or a private CBRS system from AT&T through Nokia or Ericsson. You get the benefits of security, localized data and ultra-low latency instantly sitting on an LTE second band.
"We're actually doing this now on LTE. And then depending on how much bandwidth you need and how much capacity you are going to grow to, you can then move into 5G systems probably in 2021."
Even with 5G, Boyanovsky said not all IoT applications need sub-10 millisecond latencies.
The collaboration between AT&T and IBM built on last year's announcement between the two companies that included AT&T providing IBM with some of its software-defined networking know-how. On the flip side, AT&T Business moved more of its applications to the IBM Cloud.
AT&T also has a cloud partnership in place with Microsoft for network edge computing (NEC.) The two companies' multi-year alliance included Microsoft being named as AT&T's preferred cloud provider for non-network applications as part of the telco's broader "public cloud first" strategy.
"Depending on the application that's running on the endpoint, you can be in a metro area and get really good performance and low latency at the network edge," Boyanovsky said. "The Microsoft stack will be the same proximity on our metro where our software's being distributed to get real good latency and benefits for applications to the mobile endpoint. Think of geographic-based performance improvements like gaming technology where it'd be in multi-player environments going across the consumer space as an example.
"Multi-access edge compute, or MEC, is on-prem, very industrial, very customer specific, and industry specific. It's highly secure, localized compute on-prem. So we built both. We've got a localized solution for managing an operation, and then we've got this back road type play where our hardware is turning into software and distributed closer to all the end points in order to unlock the major benefit of the application."