Transforma logo

Huawei Analyst Summit 2020: If all you have is a hammer…

MAY 29, 2020 | Matt Hatton
region: ALL vertical: ALL Edge ComputingInternet of ThingsHyperconnectivityArtificial Intelligence

Abraham Maslow, he of the hierarchy of human need, said “I suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.” It was hard to escape thinking of this while tuning in to the Huawei Analyst Summit.

Overall Huawei did a decent job of transferring what would have been an in-person event into the virtual world. I don’t want to give a full review of the whole show. You don’t have to go very far to find commentary on the various sessions on their main telecoms network and handsets businesses. Few would argue that Huawei is doing innovative things there. For us at Transforma Insights the more interesting thing is the approach they’re taking to IoT, AI, digital twin, digital transformation and so forth. There was quite a bit of commentary on all these topics.

Huawei in IoT: all about NB-IoT

It’s unsurprising that there was a lot to say on IoT connectivity. It’s the piece of the puzzle that Huawei is most associated with. In fact, when it comes to NB-IoT, which featured prominently, increasingly they’re about the only company making significant noise. It’s easy to see why. While some examples are from outside China, it’s evident that of the 100 million or so devices shipped (up 60% year-on-year), the vast majority are in China. The big apps that get name-checked are water meters, gas meters, smoke detectors and electric bikes, all weighing in with >10 million connections. The next tier are white goods, street lights, parking, connected cows, smart door locks and manhole covers (which are significant in China).

I asked about the challenges that NB-IoT has had in not delivering promised capabilities (including battery life). This has been one of the biggest challenges in pushing up deployment and use. As expected, the response was that operators need to upgrade their networks (but many, such as NTT Docomo, aren’t prepared to do that before revenue arises) but interestingly the spokesman also suggested that it’s in part a hardware issue with which Huawei needs to work with the silicon vendors. So better times could be ahead for that. The reality is, of course, that the demands for LPWA mean that the whole stack from device through to application needs to be built for low power, meaning predominantly limiting the awake time and the frequency of transmission.

The IoT track featured quite a lot on NB-IoT. It was hard to argue with the assertion that 2G apps can mostly shift to LTE or NB-IoT. Assuming the coverage is there, of course. It was, however, quite easy to argue with the assertion that NB-IoT has better coverage than 2G. There was also an unnecessary focus on better latency (10-25%) for NB-IoT. I don’t think this is a particular surprise for many, but no-one picks 2G for the low latency.

Digital transformation means 5G. Does it?

In the Digital Transformation section it was notable that the scene was set with the statement that digital transformation calls for 5G networks to serve as differentiated networks with orchestratable capabilities, dedicated networks with guaranteed data security, and self-service (DIY) networks with autonomous management. Huawei will help carriers to offer three main solutions – 5G MEC, network slicing, and 5G LAN – through their 5G Deterministic Networking driven core network. If you have a hammer, everything looks like a nail, even Digital Transformation.

Huawei IoT Cloud

Moving slightly further away from Huawei’s core business, it was interesting to hear about the Huawei IoT Cloud, its full stack offering across multiple deployment environments, aimed at ISVs/SIs. Huawei has had connectivity management capabilities in IoT for a while but this was an extension on that. Correctly it identifies that ISVs/SIs are the key route to market for much of IoT. The hosting, release, device integration and partners can all be handled through the IoT Cloud. The figures look good with 280 million connections on the platform, but most likely that is entirely achieved with Chinese operators. Not much evidence of internationalisation.

There was quite a lot of talk of digital twins, edge AI and industrial AI, but digging a little into it, the Huawei element was predominantly in providing a platform, while the actual capabilities would come from third parties. There was some reference to data cleaning at the edge, which was the third time in a week that conversation of that topic had come up for me (the others not being with Huawei) so potentially an area to watch.

A product driven approach to digital transformation

From the end user perspective Huawei emphasised that, due to the pandemic, governments and enterprises around the world have now realised that digital transformation is imperative. And, being Huawei, they emphasised that their preferred approach to these emerging opportunities would be product-driven, applying their “90,000-strong R&D team” to develop scenario-specific solutions for government and enterprise customers. Huawei has had some success with this concept in the past, with solutions that (for instance) support the multiple in-store networking needs of a retail location with a single ‘out of the box’ solution. The IoT Smart Logistics warehouse automation solution was also interesting, based on, again, experience in China, the biggest logistics market in the world.

Besides any scenario-specific initiatives, the company emphasised its commitment to a partnerships-based approach, and ability to provide first-class supporting technology. There was discussion of consolidating multiple networks to a single network, low round-trip-delays, high throughput rates, and so on. These are exactly the kinds of capabilities that you need to support digital transformation, but they are not the kinds of capabilities that necessarily help you to sell a digital transformation project. Hence the scenario-specific solution focus, and focus on partnerships.

MAY 27, 2020| Matt Hatton Previous Post
Infographic: Real world lessons in the Internet of Things
Next Post JUN 04, 2020| Matt Hatton
Talking automation, robotics and industrial IoT
All Blog Posts