AIOps is the use of artificial intelligence (AI) to enhance IT operations, using machine learning, analytics, and large data sets. Interest in the field has grown significantly over the last few years as the volume of data available to telecommunication networks and other businesses has grown exponentially. Managing such large data sets by hand is impossible, so AIOps was developed to collect data, determine patterns, diagnose trouble, and implement solutions.
I recently sat down with Aaron Boasman-Patel, the vice-president of AI and customer experience with the not-for-profit organization TM Forum. Aaron is responsible for defining and executing the strategic vision for all AI, managing across ecosystem collaboration projects, and helping set industry standards.
Why Do We Need AIOps?
At the start of our conversation, I asked Aaron to explain his definition of AIOps, and he put the focus right on automating operations. He went further by saying it’s a new state of automation where we see autonomous networks which can become self-healing, self-fulfilling, and self-assuring, and that with 30 billion devices expected to be connected to the Internet by 2030, you’re going to need solutions that can scale and really take advantage of 5G.
Aaron went further, declaring that if a telco does not invest in AI and automation, they will not have the network reliability to garner any part of the $700 billion worth of new revenues expected to be seen from the 5G B2B2X space.
Data is the New Oil
Aaron made the analogy that data is the new oil. Oil, which has existed for millennia, was basically unused until the 19th century. People didn’t know what to do with it. Once refined, it can be turned into petrol, plastics, and many other valuable items.
Data is similar in that companies have been accumulating it for years but didn’t know what to do with it. As opposed to refining oil, data has to be manipulated so that valuable information can be derived from it. We have to have the correct data in the proper format, and it has to be accessible in a safe, secure way. Then, we have to have the services that will benefit from the data analysis.
Do We Have Enough Data?
I asked Aaron if telcos had enough data to make this work, and his surprising answer was that we had more than enough to allow AIOps to improve network functionality. He mentioned one estimate that we currently only utilize and analyze 1% of all network data and wondered what would happen if we could increase that simply to 5%? It’s not a question of having the data, it’s accessing the data, and with mobile network traffic expected to generate 300 exabytes per month in 2026, there’s a lot of analysis to be done. Aaron also mentioned having toured and spoken with many CSPs, particularly in Asia, and that the amount of data they gather ranged from hardly anything to everything.
Who is Managing the Networks?
As a result of the pandemic, a great many people realized that they needed to digitize their businesses. Retail, for example, went fundamentally online as a result of COVID, and many have beefed up their in-house IT staff to manage their operations and provide the scalability needed. If CSPs don’t move quickly on this opportunity, Aaron thinks they’re going to find that people in retail, healthcare, manufacturing, and others will be managing and looking after their own networks. Aaron pointed out that JCB, a tractor manufacturer, set up an entire ecosystem to connect their heavy machinery.
Intent and Network Programmability
Programmability has meant different things to people for many years, but Aaron thinks that the next big thing in network programmability is what he refers to as ‘intent.’ By that, he meant that intent-driven autonomous networks would be able to, based on the specific service requirement such as latency and bandwidth, stack up new services in a matter of minutes, as opposed to the days, weeks, or even months it can take now.
Legalities will be a Driver for AIOps
The cloud has many forms – public, private, hybrid, edge, fog, and one role of automation will be to link all of the data sets between the different clouds, which will need AI and automation to work successfully.
One challenge will be the various legislation that will arise based on where the data is stored – on the device, near the device, in a private cloud, or elsewhere. Different rules will be regulated based on where the data is kept. AI will have to monitor the legalities of where the data is and how it can be used. TM Forum is developing an AI data governance engine designed to automate these legalities.
To learn more about AIOps, listen to the podcast hosted by Ashish Jain, CEO and Co-founder, PrivateLTEand5G.com and KAIROS Pulse. Our podcast guest is Aaron Boasman-Patel, the Vice-President of AI and customer experience with TM Forum.