Telcos need an unprecedented level of automation to stay relevant

In part two of our interview with HPE, VP and General Manager, Communications Technology Group Tom Craig tells us the number one reason telcos are motivated to adopt AI, ML and automation is to gracefully manage an ageing workforce, and that to stay relevant they’ll need more of it.

What are the practical applications of AI making a difference in the telco space right now?

Without question two areas that spring immediately to mind are network operations automation –  AI and ML will play a very significant part in that journey. Rob Harding from my team I’d describe as a world authority and how you use these technologies to drive cost from carrier infrastructures, his technology literacy is way beyond mine but he says we’re going through different phases of running networks. The analogy would be we used to use maps in the navigation world, then we had online sat nav, then we had live traffic data, then we had automatic routing with live traffic data.

Intent based orchestration, and how we use the IML in the provisioning of services, and the reconfiguration of services and carrier networks is applying those types of logics. How do we describe the intent but not hard code it in policy and scripting to allow suitable levels of AI ML to then configure the network? So in network automation, network management, and orchestration is the first ground that we’re using AI and ML technologies. And it’s a big catalyst of ‘how do you run complex networks with less people intensity?’

And my measure of success that I always used when I looked at the level of automation is you go and visit a customer operations centre in any carrier and look at the number of yellow post it notes on the service agent’s screens, and that will tell you where they are in terms of the level of automation. And sometimes you go in and its literally every inch of the screen, and that’s because  their AI and ML layer is the service agent that depths into a myriad of BSS and OSS systems to try and stitch together a product for the customer.

At one of the operators I worked with in the not too distant past, the provisioning of a VPN connection could be three to six months. And a lot of that was just the manual intensity and overhead of configuration. So I think that it’s very, very clear that you can drive scales of efficiency in the operators through the use of AI and ML technologies. For that reason it’s our number one focus area. There are already public figures that look at what the industry has forecasted in efficiency and headcount reductions they have to manage the networks with for the next five years. What’s the physical enabler? In most cases it is going to be a degree of technology driven automation that the industry has never done, and to do that the IML will come in.

There are many other areas such as fraud management –  HPE’s got a very strong position in fraud detection and revenue assurance for carriers, and again it’s an area that we use AI and ML techniques to very dynamically identify behaviours and threat opportunities for the operator. So there’s a number of very practical areas, but those two areas come immediately to mind.

In general, when industries talk about bringing in more AI applications, one of the first things that’s brought up is what this means for people’s jobs. It sounds like some of the benefits you’re describing are about efficiencies for the telcos – if they can’t necessarily achieve meteoric growth in what they are already doing, you could become more profitable by being leaner.

it’s a very it’s a very good question and it’s a question that I often ask the customers –  what’s the core driver for automation and using things like AI and ML techniques to drive a level of efficiency? And actually the number one reason they’ll say is – while the by-product is they know they need to drive a much lower unit cost to serve, operate and build new services – the number one motivator is in many cases an ageing workforce demographic.

If you look at the skills and profiles for a lot of people that have been deep into network OSS, network management, network operations, then the motivation is as that experience and the demographic of the workforce ages, how do we make sure we have a replacement technology to cater for the capacity that’s left? In the carrier world that’s far more the motivation rather than just a blunt weapon. It’s much more about how do we gracefully manage an ageing workforce demographic in network operations?

So it’s not necessarily about significantly less people working in telco in the next 20 years as all this technology comes to fruition?

Yeah and also allowing the top talent to be to be focused on the real key challenges – how do you have some of your best talent redeployed on service creation for growth? If you’re keeping the fabric running and dimensioned and capacity planned, not many customers are going to pay you additional ARPU for that, but if you can have these people opening up APIs into the edge compute infrastructures for service creation by the application developers… that vision, many operators had 15 or 20 years ago. But none have delivered it yet.

This virtualization at the edge creates an opportunity again, the operators have the highest capillarity, potentially lowest latency data centres in the world. Why not open them up to service creation. Now if AI and ML takes the mundane, routine tasks of running the network and allows you to redeploy your top talent on service creation to monetize that edge, Maybe that’s a vision which is of far more interest to all

As much as the question about jobs has to be asked, farming was done with handheld ploughs and then that involved more people than it does now – so some of this is just is just about technological progress, and the world changes?

It does but you have to go back to basics. You have to live in the commercial reality of the world we live in, and the operators… the competitive landscape is different [to what] its ever been. We had the discussion about where did all the value from the investments in 4G go? Now the top seven internet businesses have four times the market cap of the top 78 telcos. In which case, defending the status quo… I think we have to say it’s inarguable that for future viability and competitiveness, the main processes of the carrier has to have a level of automation that’s been unseen in the industry to be relevant for future service creation and for future competitiveness. It’s unavoidable, you have to have that ability.

MWC last year was awash with banners talking about zero touch operations, very few have practical examples of where they’ve done it. We’ve been working with Swisscom for a number of years, 90% of their b2b portfolio, which is a growth engine, is fully automated by HP. Their time to market for services, from order to cash has been dropped by 80%. So it can be done, and then you can redeploy people on genuine experience service creation for customers, things that will really matter and they are willing to pay for.

Check out the first part of our interview with VP and General Manager, HPE Communications Technology Group Tom Craig in which he discusses the current state of 5G, Open RAN, and how operators can evolve and grow into new business frontiers.

Get the latest news straight to your inbox. Register for the Telecoms.com newsletter here.