Verizon Making Move to AI to Address Spiraling Data Demands
Wireless carriers, like companies in every other sector, are having to learn how to use AI, said Meghna Sinha, Verizon vice president-AI, and other speakers during a Fierce Wireless virtual event Monday. Carriers don’t have a choice, Sinha said: “The sheer volume that runs through our network is so large. Detecting patterns, making sure that we can catch issues quickly, it’s just not possible through humans or through spreadsheets -- that is where AI comes in.”
Sign up for a free preview to unlock the rest of this article
Export Compliance Daily combines U.S. export control news, foreign border import regulation and policy developments into a single daily information service that reliably informs its trade professional readers about important current issues affecting their operations.
“The first thing we’re facing in this perfect storm is significantly increased network complexity,” said Justin Paul, global head-industry solutions, telecoms, at software company ServiceNow. “5G seems to be a tipping point for many service providers,” he said. With the deployment of virtual and cloud-native networks and the growth of the IoT there are "a lot of things going on in the network domain and complexity is going through the roof,” he said.
“At Verizon, we’ve been on a multiyear journey,” Sinha said. “It has become a critical part of building our networks, as we move into the 5G world,” she said: The network “has to be managed much more intelligently. Devices are constantly dealing with handoffs, from home internet to cellsites. … And then add to that all the sensors we have started to use in our daily life, but also in our working environment.” Digital twinning is also important to Verizon, she said: “We can do a lot of simulations and optimizations in a virtual environment before we do anything in the real world.”
Machine learning (ML) is one part of AI, Sinha said. “Just like humans learn, you feed a lot of data to a model and the model gets better over time,” she said. “The network has to become predictive, and it has to become self-optimized,” she said. “AI has to be at the heart of the network.”
AI allows Verizon to more rapidly deploy its 5G network and to offer customers more sophisticated solutions, Sinha said: “The devices that the healthcare industry will use, the devices that manufacturing industries will use, are more and more going to rely on 5G-type of networks. These are the sophisticated solutions where automation is going to be a critical building block.” The solutions that will run on 5G networks are very different from those that ran on 3G or 4G, she said.
Network operation centers (NOCs) were traditionally “mostly manual. That means relying on the human eye to see what’s happening in the network,” said Abhilekh Bhardwaj, head of operations-strategy at U.K. provider Three UK. “That’s changing,” he said. Carriers are eliminating manual operations by automating processes in the NOC, he said, and AI is critical.
“Technology evolution and business expansion means the technology is becoming more and more complex with each passing year,” Bhardwaj said: “I hope AI … makes the NOC more effective by hiding this complexity, thus making things happen faster.” AI can also make operations more efficient “by helping us learn and act faster,” he said. AI also will help make centers more reliable at a lower cost, he said.
“The NOCs of tomorrow will be much more distributed and will be more geographically resilient because we’ll be able to operate them from anywhere,” Bhardwaj said. “The need for people in the NOC to be in the same physical location to do their job is gone, to a large extent, if not completely,” he said.
The workforce is changing, Paul said. The younger generation of employees isn’t as willing “to do boring, swivel-chair manual tasks that could be done by a computer -- there’s an expectation of automation,” he said. Rising inflation and energy costs mean networks are more expensive to operate in 2022, he said: “If you have a large operational workforce, if they’re all expecting double-digit pay rises, that has a huge impact on your operational costs,” which only go up without new technology, he said.
Customers are also more demanding, Paul said. Customers look at the big hyperscalers, software companies and companies like Apple and Google “and say I want that experience when I deal with the telco operator,” he said. “Changing customer expectations means there’s an expectation of a much quicker approach to things,” he said. Carriers have to play catch up and networks are probably less than half automated, Paul said. But the era of AI and ML operations is just starting and “we’re a long, long way from the concept of zero-touch operations,” he said.
Operators in the future will be able to offer AI services , tied to their networks, said Joe Krystofik, Fujitsu head-product planning. He cited as potential business cases applications like traffic-control systems, robotic warehouses and remote-access monitoring. “Medical centers may need a private network slice specifically dedicated to doing their operations,” he said. AI will help carriers “orchestrate, manage and provision … and monitor” that slice, he said.