Connectivity is everywhere so it should be simple to specify the connections you need for your IoT deployment. However, the decision process remains complicated and fragmented, reflecting the complexity of IoT deployments themselves, writes George Malim.
The IoT connectivity challenge now very much resembles The Rime of the Ancient Mariner. It’s a case of there being water everywhere but never a drop to drink – or, in this case, connectivity everywhere but not the type your application needs. This is because of the difficulty in finding the sweet spot in the IoT connectivity equation. There are three core elements to this formula: availability, performance and cost to which I’d add a set of nice-to-haves including: simplicity, service, scalability and flexibility.
It’s easy to find available connectivity. Simply talk to a satellite provider and they’ll sell you coverage that encompasses the entire planet. Vol. 2, No.1 I IoT Connectivity I 05 It’s easy to find cheap connectivity. Speak to low power wide area network (LPWAN) providers or the owners of lower bandwidth cellular networks and they’ll sell you connectivity for cents per month per device.
It’s easy to find low latency, high capacity connectivity. Sign up with your local 5G provider and access speeds of more than 10 gigabits per second with ultra-low latency, sufficient to support high definition gaming and beyond. The problem is, you can’t get all three attributes at the same time, let alone from the same provider globally. There are only a few applications today that demand high capacity, minimal pricing and global footprint. For most, the business case either demands low cost and can accept limited capabilities or can sustain higher costs in order to gain the throughput the app requires. In addition, today there are few truly global deployments so covering the major markets is sufficient. However, we are still only at the dawn of IoT. The massive upswing into the tens of billions of devices has not yet arrived and this will necessitate a global footprint. The large cellular carriers and a number of connectivity providers are moving in this direction but the idea of a global network that is owned and operated by a single entity is unlikely to come to fruition. Instead, carrier groups will collaborate and IoT connectivity specialists will form relationships with a limited number of partners to ensure global connectivity availability for their customers.
Enterprises can simplify their connectivity selection by referring to what they already know. If you provide soil monitoring technology, you know you need rural coverage, cheap, easy-to-deploy devices and cost effective access to a network that offers relatively low bandwidth so your devices can report small amounts of data at a fixed regularity. LPWAN offerings are probably the best fit here. At the other extreme, if you’re providing public transport data to a smart city you need, faster speed, continuous, qualityassured connectivity that has good urban capacity. 5G is a good choice here.
If you’re launching a new service the things you know you don’t know need to be taken into account. These range from the basic such as: Will customers like the service? to the complex such as: Can I get coverage to support the service in Mongolia? Enterprises recognise that until an app is out in the market, it’s hard to estimate how many connections will be required. LTE-M might be cost-effective for a trial involving 3,000 users but, if the app takes off, the numbers might not work for one million users. Narrowband IoT (NB-IoT) might be a better bet. Equally, a successful global service will need to extend into new markets. An enterprise that has used Vodafone, for example, for its traditional enterprise communications needs and uses it for IoT enablement may find everything works well in countries where Vodafone has its own operations. It may also find that in countries where Vodafone has strong partners, the connectivity performs as expected and the cost is sustainable. However, as the service extends into other locations, the original provider may not be able to provide service and the enterprise will have to find and manage new partners. Making decisions that provide flexibility to address the things you already know is straightforward. It’s also relatively simple to select providers in preparation for success of an app and to predict which geographical areas will need coverage first. Many connectivity providers are working across IoT and have experience to draw on from previous deployments and will be able to help with strategies for scaling up or down and extending into new markets. There is, however, a further layer to consider. What about the unknownunknowns?
You can’t provision connectivity for a situation you have no awareness of. That’s scary but true. Today’s app that is ticking over at 10,000 users may take-off or it may be made obsolete by a new innovation. For long-term deployments of more than a couple of years, this is a big issue because it means enterprises are trying to dimension connectivity supply for way into the next technology cycle. If, for example, you’re a car maker, your vehicles will be in use for at least ten years. Equipping them with LTE-M is good for today and 5G is good for the next five years but what will the next apps be. Will they require more capacity than 5G, how will that be retrofitted to vehicles and how will you make money from providing it are all unknowns. It’s abundantly clear that good technical options exist at either end of the spectrum, no pun intended. There are low cost, low capacity solutions that are simple to deploy and are set to offer great geographic coverage. There are expansive, resilient, high capacity options that have probably more limited coverage. In addition, there are midmarket solutions that do most of what you need, most of the time. That, perhaps, is the goal: to select connectivity that has most of the functionality your app needs, more of the time than alternatives. If you can find a solution that does this for 80% of your customers, you can address the more difficult 20% with alternatives and do so in ways that are scalable and resilient. No one said this was going to be easy but this issue of Tech Trends provides substantial detail on the various options and provides information on how to plan your connectivity, blending different technologies from different providers to give your deployments the connected foundation they need to thrive.