In the run up to MWC, Deutsche Telekom has unveiled nuSIM, a new system that provides embedded SIM (eSIM) functions at a price-point suited for low-cost IoT devices. With a slew of partners onboard, the news comes as the MNO tries to position LTE as its primary LPWAN technology for long-life applications, to fend off the incursion from cheaper U-LPWAN options.
Notably, no actually prices are discussed in the announcement, nor is a gesture towards the potential reduction in the cost of such modules. What is mentioned is the type of devices that DT anticipates making use of the technology, which are ‘mobile IoT applications with a long-life span, such as asset trackers or smart motion or temperature sensors.’
Now, unless MNOs like DT decide to radically shake-up how they go about pricing for its network connectivity services, this does seem like a pretty damp squib. If the cost of an L-LPWAN module clocks in at around $10, even if the nuSIM optimizations represent a 50% saving on this bill, that $5 is probably good for around a year’s worth of connectivity, should you strike a great deal with your MNO.
To put it another way, the much greater cost for L-LPWAN devices in this current climate is the network, rather than the hardware. As such, paying even $0.50 per month, a bargain for most cellular carriers, rather defeats the potential cost savings brought about by nuSIM, if the ‘long life-span’ is more than a year. When we talk LPWAN devices, we are usually envisioning the 10-year gold standard for battery life, and so things become a bit confusing when arguing in favor of things like nuSIM.
So, it’s not entirely fair to compare this with a U-LPWAN technology like LoRa, as the applications can vary quite significantly. The need to have national network coverage to support the mobility part of the application muddies the waters in these sorts of comparisons, but as it stands, there’s a looming inflection point where the cost of L-LPWAN hardware ends up being cheaper than U-LPWAN, thanks to the economies of scale at play.
LoRa has a big advantage for campus-based projects, as the user can install a smattering of base stations and cover an entire university, hospital, or business park for a fraction of the cost that they would incur if they had to pay per-device monthly fees for all those devices to be connected to an MNO’s L-LPWAN network.
But of course, the MNO’s should have an advantage in the coverage footprint side of things, if their hype about the ease-of-upgrade to L-LPWAN is to be believed. Things have improved in that respect, and while LoRa and Sigfox are both surging forwards, the MNO’s do seem to have an advantage in selling a customer the vision of guaranteed coverage.
Of course, that sort of guarantee comes at a price, and so we have circled back to the initial problem we have with projects like nuSIM – that this saving is good for only a year or two’s connectivity fees. If an application is very important to a user, than they are willing to pay more for it, and so saving a few dollars on a $1,000 device with a $30 monthly fee is pretty negligible.
The marketing image that arrived with nuSIM features a delivery truck, a factory, a robotic arm, and a conveyor belt that could conceivably be in a distribution center. We don’t think any of those are strong candidates for the kind of low-power and therefore low-value applications in which the BOM savings brought by nuSIM might be just enough to greenlight a large deployment. Those are applications that would be served by the higher-power LTE variants, surely, and so they are in a different pricing tier.
To be clear, nuSIM is a good thing if it reduces the cost of a given device – it just seems to be being positioned strangely, given that the current largest concern for a developer opting for a licensed cellular connection is the fact that the network costs can easily dwarf the total cost of the hardware involved in the device, over the desired life of the product in the field.
However, the more radical step that would certainly change our opinion greatly would be a change to how L-LPWAN connectivity is actually billed – moving from a per-month model to one that actually charges per-message. This is something that Vodafone has floated in the past, and it’s an approach that makes a lot of sense for the applications that DT has name-checked in the nuSIM announcement itself.
This could enable asset trackers to send messages only when they are woken by movement in the thing they are attached to, or for temperature sensors to only send a message when a temperature threshold has been exceeded, or if a leak has been sensed. For many applications, this means that the per-month message threshold might be as few a handful of messages – and if a low per-message price is set, then suddenly the cost savings of nuSIM actually make some sense to decision makers.
Suddenly, the kinds of devices that only need to send messages on a very occasional basis, basically the weekly ‘I’m still alive’ message and whatever critical or warning payload they’ve been allocated to handle, are now very viable, as the network connectivity fees for the five years until the battery dies might easily come in at under $5.
At this point, the nuSIM savings might mean that a developer opts for LTE, rather than LoRa or Sigfox. That change in dynamic could really steal the lunches of many U-LPWAN network operators. However, there are so many variables at play in the decision-making chain that generalizing as we have can make things seem much more black or white than they actually are. There is no typical customer in the IoT.
DT has racked up the support of most of the suppliers for the L-LPWAN silicon. For chipsets, it has announced the support of Altair, Commsolid (Goodix), HiSilicon (Huawei), Nordic Semiconductor, Qualcomm, Samsung, and Sequans. For modules, Quectel, Sierra Wireless, Telit, and u-blox are on board, as is Nordic. And finally, for the digital security aspect, Giesecke+Devrient (G+D) is the provider. Gemalto and Idemia are noticeably absent from that list, and this looks like a pretty good win for G+D, if it really is the only SIM-provider involved.
DT says that it is “ready to share the technology with other operators and vendors in the IoT segments, to establish a viable ecosystem. Interested parties are welcome to join.” Crucially, it appears that nuSIM is an open specification, meaning that it has a pretty good chance of being adopted by other MNOs, and therefore not confined to global irrelevance.