In Part I of my MWC19 review, I covered the changes to industrial structure, but no review of MWC is complete without one or two tech highlights from the telco side – one of my main professional loves: –
Telco Cloud – finally seeing some statements of the gob-smackingly obvious emerging around the operator cloud message – with Orange and Vodafone announcing that they are creating a joint approach to NFVi – it seems operators finally understood that in order to achieve the same economies of scale as hyperscalers they cannot act as individual operators when defining infrastructure requirements – they need the scale of the industry. It’s taken a long while for this penny to drop and hopefully this announcement triggers others to think this through – hopefully before 2030.
Far more exciting news in the telco cloud space, however, was the announcement of a Cloud Native operator being created in Japan by Rakuten and an interesting mix of vendors. This is a definite game changer if it is successful and is one to watch closely as they also decided they needed to develop some of their own hardware in order to do it – this is much closer to a true hyperscaler approach and its launch in October proves to be a highly interesting test of cloud native in telco.
Outside of telco cloud, IBM and Vodafone announced their joint initiative for Enterprise cloud – that seems like an interesting combination of skills and again is something to watch with interest.
5G was less clear and really, I just left with more questions than I arrived with. 4G made perfect sense to me, whereas with 5G I am left with more than one question mark. I share them here in the hopes that someone reads them and at least starts asking them in their own companies. I welcome any feedback on any areas.
5G seems to be full of many new buzz words like ultra-low latency and enhanced mobile broadband along with NR and network slicing. All of this sounds very exciting – what I am missing is a real use case. One where I cannot do it without 5G? With 4G, both the demand and need were made clear by 3G implementations – I am struggling to see this for 5G.
Greater flexibility in the radio gives the opportunity to more effectively manage resources and network slicing allows greater application flexibility and certainly one can imagine many industrial use cases for such concepts – but I am left wondering how the business model really truly stacks up? Massive machine type communications – as mentioned above – are starting to see business models that work. But which ones of these really require 5G and can’t live on 4G? As discussed in CIC’s Briefing Note: Could orchestration and automation be the blockchain ROI link for Telecoms), these types of 5G use cases imply a completely new manner in which operators need to coordinate with their customers and will need new solutions to manage across enterprise boundaries on behalf of their customers as a result. This is a completely new industrial logic for the mobile industry, and I left wondering how long it will take to see real profitability on 5G networks while operators attempt to work within it. Currently I see many a hole in many a business model. Not at all insurmountable but full of holes nonetheless.
As mentioned above, 5G really comes down to new partnerships and a new understanding of industry roadmaps. There is a technical question emerging, however, – what is the right balance of coverage for a nation to get good ROI for operators on 5G investments? If the majority of the use cases are industrial that implies a different set of customer relationships – but also a different spread of coverage – and base stations – for applications. Industrial areas are not the areas operators usually cover first – and we can see in many European installations that they plan to cover end-user consumers first in large cities. How do you then ensure the industrial use cases can be met? Autonomous vehicles for e.g. trucks will require national coverage. What’s the right balance of spectrum then? What partnerships are required to deliver those at telecom scale? These are just some of the questions I left with; to me it implies completely new investment models for the industry – in particular for spectrum management and sharing. Telecommunications is run by a massive 10-year investment cycle – does that mode of operation still work if networks are meant to be sliced? Data sharing of the nature suggested by many of these use cases also requires investment models that currently don’t exist. The industry therefore needs fiscal innovation as much as it needs technical innovation for 5G to fly.
Naturally, I can’t do a summary of MWC without discussing blockchain. Many people know I have been working in blockchain and telecom for quite a few years now – driving a discussion about how it can be used effectively in different parts of the industry for managing the emerging technical and business challenges in the industry (listen to some of the podcasts on the topic done by myself and others in the CIC team at MWC). I am excited about some of the work done by start-ups including the Bristol-based Zeetta Networks and also the announcement of the Hyperledger Telecom SIG – if blockchain will take off in telco, it needs to be vendor neutral in order to allow for greatest flexibility across the world. More importantly, however is the understanding about the requirements for carrier-grade blockchain solutions – which is about more than just ensuring a consensus protocol is robust – it is about security, business processes, structured network management. In addition, it is about well-trained developers and system administrators – this is something that the upcoming Hyperledger Certification and IBM’s Composer Suite is helping to deliver.
For those looking to Blockchain phones – I am cynical. Until you work out how to share spectrum using them, this is really just a wallet on a phone. Not even close to a breakthrough, but a handy app.
The Upside – there’s a lot more life in this industry yet
The really exciting news for me was that IoT has started to move out of PowerPoint and into our lives. The collision of ICT and Operation Technologies (OT) has finally started to find its footing and its business model – some of the most exciting things we saw in this area were the AT&T IoT data being linked directly into the CRM solutions of Salesforce. While at first glance this might look quite sedate – the power of connecting IoT data directly into a customer business process flow opens up numerous opportunities that have been lacking before – both for industrial end-users but also for smart city installations. It can provide the missing link between infrastructure investment in IoT and the revenue streams associated with them.
This type of technology innovation allows for the effective management across transaction boundaries in a way that both companies can benefit from. There were many other solutions like this out there and it finally seems there is a way to link all the great work presented at 4YFN (4 years from now) into the large-scale industrial structure of the telecom industry.
The other surprising but positive insight was a number of smaller companies providing exciting solutions that hint at a much broader role for operators if they can seize a new mentality and way of thinking about their business. The first one was VNC automotive who have repurposed the old VNC software (remote desktop for those old enough to remember it) for the automotive market. Using the pixel shifting software, they are able to offer a pretty neat autonomous vehicle experience – using 3G if needed. In the old days of software, we were forced to think deeply and ‘cheaply’ about our bandwidth – in the days when it cost a lot of money to transfer a Mb, we were thrifty in our system designs as a result. This VNC solution to me represented a truly innovative approach to the question of autonomous vehicles and answered the question for me – “ok, but when will we have national coverage of 5G?”
One of my favourite new finds at MWC was a small company called Saguna who provide some extremely nice and very usable cases for Edge / MEC over 4G networks – and presumably excellent for 5G as well when it arrives. The use case that really got me thinking in a new ways was the one where they are able to provide hybrid public/private access across the same network – you can imagine many uses in manufacturing or other types of environments where such a concept would be useful.
In short – despite the industrial issues I outlined in Part I – there’s a lot more life in the mobile industry yet and we are in for some exciting moves over the next few years.