In part one of this Telecoms.com interview with HPE, we talk to VP and General Manager, Communications Technology Group Tom Craig about the current state of 5G, Open RAN, and how operators can evolve and grow into new business frontiers.
You launched the HP 5G labs a couple of years ago now. What is being worked on there and what has already been launched from it?
We’ve got a 5G lab in Fort Worth, one of the things that we’re working on is definitely wifi and 5G interoperability. The customers want a best connected network strategy, wifi and 5G is going to coexist so how do we make sure that we have policy, security, and handover between the technologies in a very graceful way?
It’s also been an ecosystem where we’re very open to inviting industry partners to come in and collaborate and co-innovate with us. Whether its manufacturing, zero defect manufacturing, optical inspection, manufacturing using 5G, healthcare, oil and gas and a number of other areas. We’re working in piloting industry vertical solutions, and fast prototyping them, or proof of concepting them in the lab, and then taking them into production in the field.
We’ve seen a lot of interest from retail, manufacturing, and health care and also defence, which has become increasingly a very interested party looking at 5G deployment services. So in the lab we have partnered with a number of ecosystem partners to prototype industry vertical solutions, and then very often build and take to market.
What sort of applications is the defence industry looking to develop with 5G?
A lot of them require very highly secure, ultra-low latency… very secure network slicing technologies of an automated nature. And very easy to deploy. At the moment we’ve got some ruggedized solutions that defence are interested in, where you can spin up a 5G network in a matter of hours and put it into hostile environments, which is unfortunately the need in the case. So for defence we have seen very strong interest and are actually in trials with a number of military applications today, which obviously for security reasons we can’t go into in great detail – but it’s typically characterised by being ruggedized, easy to deploy, self contained, and spun up very quickly and easily.
You’ve mention a few B2B and industrial examples there, is that where all the innovations and applications are with 5G?
There are other consumer applications that are talked about – gaming in augmented reality applications – but the difficulty is so far none have passed the acid test that I’ve seen in terms of deployment of and ‘will the customer pay more for this?’ In the B2B space though there are very clear areas… in factory optical defect inspection for zero defect manufacturing, the return on the business case can be profound. And so will customers pay incrementally for the solutions? Absolutely.
There are other [5G] consumer applications that are talked about – gaming in augmented reality applications – but the difficulty is so far none have passed the acid test that I’ve seen in terms of deployment of and ‘will the customer pay more for this?’
Likewise, in healthcare, likewise in retail and other areas, there are very clear use cases. They’re never connectivity use cases, which is mindset shift we’re going to have to work with the operators on. In most of the use cases it requires quite a depth of digging into specific industry business processes, to really be clear on the value of the application that you’re going to deliver. And also you’re kind of seeing a whole breed of industry specialist SIs is in each of those niches that are bringing some of the creativity to market also.
I suppose we’re still waiting for the killer app in the consumer space.
I think it’s more difficult, I’m not saying it won’t come, but in the B2B space the business models for how you can incrementally monetise are much clearer, and that’s why a lot of the focus is on those areas. Whether it’s in smart cities manufacturing, defence, healthcare, there are very clear applications. In terms common traits, some of them are in challenging radio frequency environments – oil rigs, manufacturing, seismology, mining – often dispersed environments or factory environments, but particularly where you need predictable, high mobility within that environment. When you see kind of some of the mobile robot technologies that have been used in hyper warehousing estates, then then you need that ability to have very predictable mobility and roaming within the radio coverage. Then we start to see some 5G applications really come into their own.
How big a part do you see HPE taking in the Open RAN space, and how do you see its trajectory changing over the next few years?
There are often academic discussions around ‘will it, won’t it happen?’ – in my mind its unambiguous, its only a matter of timing. The same debates were had over network function virtualization and now it’s just in core networks, and there’s now not a debate. It’s happened. It’s de facto, it’s what you do in terms of building network infrastructure. Our view is it will happen, the question is… it’s a timing and it’s a journey. But step one, you need to make sure your architectures and designs are building a way to get there. Step one – virtualization at the radio access edge – we think is well is well underway, and we’ve got engagements with a number of operators around the world.
Going back to kind of one of our core assets, The biggest vRAN deployment in the world today is with a US tier one operator, the automation and orchestration elements of that deployment were delivered by HPE, which we’re very proud of. I think that’s come with a lot of learnings, which we then look at how we productize. We then look at some of the product announcements we’ve done. We’ve talked about our RAN automation software capabilities, using some of the learnings and knowledge of that journey, and we’re then using that for commercial deployment of cloud native software for the automated deployment of compute infrastructure at the edge.
There are often academic discussions around ‘will it, won’t it happen?’ – in my mind its unambiguous, its only a matter of timing.
That will automate all the tasks of firmware, BIOS security, configuration, staging, all of that stuff will be done from the cloud automated by HP software capability. Right up to the software management orchestration at the RAN edge, that will also be automated by software capability by HP. And not just for 5G, but also retrospectively for 3G, 4G and 5G networks, which not much of industry talks about. If you only solve the problem for virtualization at the edge and you only talk about 5G, then what about the other 90% plus that keeps the lights on? We think in the next 12 months we’ll start to see a growth on the virtualization phase all over the world, through Europe, India, all the way all the way through the world and that’s the big focus we’re on now.
There’s a trend where telcos are increasingly moving over parts of their business to the public cloud through hyperscalers, whether its AWS, Microsoft Azure or Google Cloud. Do you think there are any downsides to doing that?
What we’ve seen so far is that for the operators, the adoption has been to address and service their needs for their internal core IT. If you think of three kinds of different envelopes, there’s the core IT element, there’s the technology for core network, and then there’s edge network. The adoption that’s been seen – and it’s driven agility and some good economics for some of the operators – is in the IT space.
But when it comes to core network build, particularly on edge network build, then I think public cloud has a number of limitations. If you look at some of the operators are looking at the future of the world, they say actually this is the foundations of the house, and it will also be the foundations for new service creation. And new services, whether it’s mobile edge compute solutions or others, the vision that they have is how do we open up this infrastructure as a service creation and innovation platform? And as such, we need to own build, deliver and operate.
70% of data is not in the public cloud, it’s at the edge. And actually, it’s growing and growing rapidly. And innovation is growing rapidly the edge. From my 31 years in the industry, I think that service creation and innovation at the carrier edge is the lifeblood future of new service creation and growth opportunities for operators. So to defy the forecasts of 1% service revenue growth that some of the analysts are coming out with for the next five years, one of the opportunities is definitely to take the core asset the operators have, which is – if you look at their local exchanges central offices – they have a data centre estate that has the highest capillarity and potential lowest latency of anyone on the planet.
Go back to the macro economics of the industry – anaemic growth, huge investment hurdles to build the 5G infrastructures – then what’s left for genuinely new service creation and innovation?
So in many cases, they can be the physical manifestation of the edge – but only if it’s theirs. If you surrender that edge… you get operators who’ve made solutions and just resold some public cloud offers, they get into these philosophical debates about whose customer is it? Where’s the value creation? Who’s the contracting party? My view and HPE’s view is that unambiguously it is the carrier’s customer, it’s their edge, and we can help them build and innovate on it for future service creation. Without question public cloud has played a strong role in some of the economics of delivering internal IT for the carriers. But when you come into core network and particularly edge, then I think that’s an area of service creation that’s needed by the operators.
Then the question is what role can HPE play. And one of the reasons I joined the company was fundamentally the positioning of Greenlake. When I think on it from my operator perspective, the idea of having cloud experience, cloud economics – matching cost to revenue and income, or pay as you scale, which is the business model of the public cloud – having that but having the agility and ownership of the edge so I can match income and new service creation with cost, I think is a very attractive proposition for the sector.
Even in my time here, there’s a number of cases where operators have been able to take new services to market, such as KDDI in Japan. We launched an MVNO with them entirely on Greenlake. But the key difference being we’re not going to sell you a tonne of things, we’re going to build together the MVNO platform, so you get the agility and the controllability of the cloud experience but critically, they can match cost and income, which is a big, big, big issue for the sector.
When I talk to people at MWC and on one to one customer visits, one of the biggest challenges is Capex for investment and service creation, no question. Go back to the macro economics of the industry – anaemic growth, huge investment hurdles to build the 5G infrastructures – then what’s left for genuinely new service creation and innovation?
Being able to say to industry with Greenlake you can take the new offer to market, whether it’s to consumers or to the b2b segment, but we will jointly match cost and income to grow together, I think is very powerful. Oh, and by the way, it’s unambiguously your customers. I think that’s something which resonates very well with the sector, and I think its going to be an increasing trend.
From your perspective do telcos have the right business model for growth in the medium to long term? Are some of them confused about what that business model is, and what should they be doing that they aren’t?
I’m very optimistic about the opportunities in the telco sector for growth. I think consumer can remain strong, but I think that the B2B sector in particular has got very strong opportunities for growth for the sector. in one of my previous roles when I looked at the strategy for growth, you could see that if you’re only addressing the kind of the core B2B mobile market, you can legitimately describe very quickly an addressable market 3, 4, 5 times the addressable market value by just addressing clear adjacencies. And by clear adjacencies let’s just take a mobile only operator, there’s a diversification play into fixed communications, whether that’s software defined networks or others.
Then you’ve got technologies go mobile, fixed and cloud based telephony – which is a pivot a lot of them did some time ago. So you start potentially cannibalising the old customer premises equipment market that was doing telephone switches or call centres. You can then legitimately move into the office automation software suite, because if you’re doing subscription basis mobile communications, fixed communications, cloud based telephony, you can then move into real time and non-real time communications.
Fixed, mobile, cloud telephony, cloud contact centres, office automation, payroll, finance – there’s a whole branch of the world where you can easily be seen as a legitimate broker.
So a lot of them then said, we can also legitimately subscribe to office automation like Office 365 or SAP… you can describe a journey where you can easily become the vision of the cloud IT manager for business. And lower latency, high availability networks can be a massive legitimising factor on that journey. Fixed, mobile, cloud telephony, cloud contact centres, office automation, payroll, finance – there’s a whole branch of the world where you can easily be seen as a legitimate broker. But then you have to make sure that you’re able to look and act like a digital native to the customers on that journey so that each step is super low friction, that it becomes a very continuous journey. And you get to this world of subscription based communication and IT services for business, and some operators are already marching down that journey and showing good growth. There is a very, very clear path you can drive for growth and there are industry case studies that have done it very successfully.