I was browsing through LinkedIn this morning when I came upon a post asking a simple but very relevant question for anyone interested in the evolution of the Internet of Things:
NB-IoT or LoRa ?
Personally I think LoRa and NB-IoT will coexist for a while. Mobile network operators (NB-IoT) in North-America are a lot more excited by 5G right now. Also, they just don’t have the agility to be successful in such an early-stage effervescent emerging market.
This is why smaller outfits such as X-TELIA will pave the way, exactly as in 1992-1995 when small Internet Service Providers (ISPs) created a new market thanks to disruptive technology. When it matured, carriers eventually got in the game and rolled up that market.
IoT is more complex than was dialup Internet. Every solution is a different set of HW/SW, and data is useless unless it’s actionable. Telcos will soon offer NB-IoT connectivity but will struggle to deploy custom solutions which is not part of their DNA.
One more thing. The 3GPP carrier ecosystem (NB-IoT) is heavily taxed by a multitude of patents which will make it very difficult to bring to market sensors and solutions able to compete on price with LoRa or SigFox equivalents. I am talking here about the Massive IoT market: A very large number of very inexpensive battery-operated sensors transmitting simple data for many years across long distances. According to Ericsson’s November 2016 mobility report (and half the tech analysts on the planet), Wide Area IoT will experience a 30% CAGR between now and 2022. I guess that’s how you get to 30-50 billion connected devices worldwide by 2025…
Now getting back to our question, personally I think there be coexistence, with NB-IoT solutions running on dedicated private carrier spectrum addressing more critical applications, and less critical LoRaWAN applications running on license-free ISM bands.
Let me give you two examples to illustrate this:
LoRaWAN: If you have hundreds of humidity sensors in a farmer’s field sending regular updates, it’s really not the end of the world if there is a problem getting one sensor’s data through at a given moment because another sensor transmitted on the same frequency at the exact same time, creating interference. That’s a perfect application for LoRa, running on the public, free ISM band. Cheap. Powerful. Ideal for applications that are not critically time sensitive and where high value is derived by analyzing data from many sensors. Think of all the Smart City Applications for which LoRaWAN is a perfect fit.
NB-IoT: Although today many portable credit payment machines operate on 3G/4G/LTE in many parts of the world, their operating cost would drop substantially if they were migrated to NB-IoT. Using 3G/4G/LTE to process such a small amount of data as a credit card transaction is like using a bazooka to swat a fly… Seriously. You can watch an HD movie on a tablet in your car zipping down the highway with that kind of bandwidth (assuming someone else is at the wheel…). Not only is it totally overkill from a bandwidth perspective, but the more bandwidth a wireless technology provides, the more energy it needs. Since every credit card transaction needs to happen in real time and is critically important, using the carrier’s private spectrum with NB-IoT is a perfect fit. In this case relatively high value is derived from the data of a every single transaction of every single single connected device. That’s an example of the Critical IoT market.
It looks like I am of the same opinion as many European mobile operators such as Orange and Bouygues Telecom who are currently deploying nation-wide LoRaWAN networks which will co-exist with their 3GPP wireless networks. I wouldn’t be surprised to see them eventually launch a private-spectrum-based NB-IoT offering alongside their new LoRaWAN networks as there will be a market for both, in perfect co-existence. We shall see…
All in all, a very exciting time to be in the IoT business !
Eric Bourbeau, B. Eng, MBA, Founder and CEO of X-TELIA Group. firstname.lastname@example.org