Capitalising on IoT
IoT is happening now. Some players will be restricted to niche markets, some will win big, time will tell. They key is to find a way to make business!
Many operators are looking into IoT as source for growth, but I suspect major share of them don’t yet know what technology to choose. Most will probably automatically lean towards 3GPP while some will choose another path.
During my previous projects, I’ve had many interesting discussions about IoT, and I thought it would be a good idea to tie a few things together into a blog post. I will look at this from a few angles: What is going on in the industry now? What are the key features needed in IoT? What would be a viable go-to-market strategy?
On industry level there is a clear recognition of potentials for IoT but what it means as true business potential is still to be seen. In any case, the industry is moving fast and many things have happened recently. These are the most well-known standards:
1. NB-IOT Two competing cellular standards (NB-LTE and CIoT) for IoT was merged around September/October 2015 to be called NB-IOT (Narrow Band Internet of Things). The target is to have a formal standard by Release 13.
2. LoRa LoRa standard currently utilises unlicensed spectrum and is supported by the LoRa Alliance. At least Orange France some time last year picked standard for deployment. At that time there were still two separate cellular standards for IoT. While the use cases for LoRa may be different to NB-IOT, I wonder if they will revise the strategy, now that NB-IOT is almost definitely becoming the dominant design.
3. SigFox SigFox – the French IoT startup, which no longer can be described as startup – builds on their own proprietary tech. SigFox business model is to build their own IoT-dedicated network and to supply the devices through distribution partners. Their objective is become a global IoT operator and the de-facto standard. Sigfox seems to be getting at least some traction in Europe and US. As of now they claim to have penetrated 14 countries with sites covering 1.2 million square kilometres and 223 million people.
4. LTE-M (3GPP) There is the *slightly* older (or more accurately, longer ongoing development of) 3GPP standard LTE-M (for Machine-to-Machine) and LTE Category-0 terminals, but that is deemed to be too high bandwidth (and hence too restrictive link budget) for IoT purpose.
And Competing Business Models
Which one (if any) of the four will eventually win the race? I suspect it is not the technical details, but it’s the business model (and its suitability to the local and global market) that will play a bigger role in driving the adoption. And the business models do differ:
- NB-IOT / LTE-M: Facilitate mobile operator IoT/M2M business with existing network infrastructure
- LoRa: Develop an appealing technology and lure the IoT industry to adopt it
- SigFox: Become a global IoT operator
One thing is clear about the “chicken vs egg” question: in this case the network and coverage must be available first, then devices and applications will come. It will not happen the other way around. This leads to another topic of “killer-app” for IoT: I don’t believe there will be a single killer-app for IoT but rather IoT will become a platform that allows the hundreds of thousands of apps that will flourish in their own ways.
A Few thoughts about NB-IOT
Features I expect to see in NB-IOT (without looking into too much details of the ongoing Working Group progress, but simply thinking out-loud about what makes sense to attract operators to utilise NB-IOT):
- Co-existence with LTE (and maybe GSM), with some substantial differences like removal of mobility function (Electricity meters, fridges and light switches don’t typically move); and total revamp of the mechanisms related to paging and tracking area updates with target to save battery life (target battery timespan for IoT devices is 10+ years)
- Live inter-operability with LTE (and maybe GSM) – ie the network will reconfigure itself to dynamically made available for NB-IOT when demand arrives (similar to eMBMS)
- Software based implementation at base stations, with no/minimal hardware needed assuming same LTE spectrum will be used
The business target for the technical specification would therefore take care of the majority of users and use cases of IoT, but at the same time somewhat conflict with that of “traditional” M2M.
On high level, what operators need to do to bring this to market:
- Select the technology (or technologies)
- Activate technologies on the sites (NB-IOT/LTE-M/Cat-0) or rollout new network/overlay (LoRa/SigFox)
- Build the back-end/middle-ware, or partner with someone with that capability
- Educate the market and sell the service
And regarding the 3rd point, it’s imperative to understand that IoT is all about integration – this is possibly the most important bit: IoT cannot be sold as simple connectivity. Instead, M2M/IoT needs to be sold as a service (APIs, device management, security, etc and also typical BSS stuff like billing, activation with concentration on self-service).
Some quick examples from EverythingEverywhere, Tele2 and another one about Tele2. You can see both EE and Tele2 have partnered with the same bunch to deliver the services. The key is to reduce the barrier of adoption so that SMEs/general-public with minimal technology knowledge can understand and start using it.
I think we can all agree the IoT is happening now. Some players will be restricted to niche markets, some will win big: time will tell. In order to not to miss the train, it’s important for mobile operators to soon start:
- Technology roadmap/selection process – driven by addressable market of different applications enabled by different technologies
- Rollout planning – with different link budget there would be some work related to coverage planning no matter which technology is chosen
- Defining the process flow for the middle-ware/etc to work – though I see that as more of role for BSS vendors/consultants
All these developments are fun and exciting – I look forward to getting involved in the near future!