Finding the value in satellite IoT
The global demand for satellite capacity is forecast to show a compound annual growth rate of at least 30 per cent between now and 2025. Most of this will result in greater data capacity which will be allocated to consumer broadband and backhaul segments where growth in traffic is expected to outstrip the inevitable year-on-year declines in 'per MB' pricing for satellite connectivity. This has already seen operators engage in a fierce race to serve a burgeoning market with their next-generation high-throughput satellites (HTS).
Over the same period, many IoT applications will have firmly emerged from the trough of their hype cycles, as Gartner would say. Perhaps this presents the satcom sector with a complementary growth opportunity which is immune from the laws of raw data capacity supply and demand.
In absolute terms, it would be a mistake to assume IoT will pose any significant capacity challenge for the satellite operators. Many of the predicted 50 billion IoT connections are each expected to consume less than 0.5MB per month – most of them perhaps using significantly less.
Putting these numbers in context, a mere 10Gbps continuous capacity (less than one per cent of the total capacity promised by OneWeb) might be sufficient to serve the demands of three billion asset tracking devices – or 600 billion smart meters.
Admittedly, connected 'things' are not uniformly distributed around the globe and there may be a second-order consideration to capacity density requirements. Unlike broadband, the major concern for satellite IoT strategists is less about ensuring adequate capacity – build it and they will come – and more a question of how to monetise IoT at all.
Whether they are built on a satellite or a terrestrial communications platform, many IoT opportunities present significant challenges to the connectivity supply chain, given that:
- Connected 'things' tend to generate extremely low volumes of data to bill against
- The value contained within each unit of data varies massively across different IoT applications
Unfortunately, the value delivered by an IoT system is only loosely correlated with the volume of data it produces. By way of example, a single intruder alert message sent just once by a perimeter fence carries a very high value. Whereas, a thousand similarly-sized messages, sent by smart-city footfall sensors carry significantly less.
This might sound obvious, but it matters.
The volatile relationship between the value of connecting a 'thing', and the volume of data it produces, means that conventional 'per MB' monetisation plans cannot address IoT. Not by a long shot.
'Per message' monetisation is a step in the right direction but it is still hard to see how uniform pricing-strategies can ever be optimised across a broad superset of IoT applications. Nonetheless, terrestrial LPWAN operators seem to be aggressively averaging-down their pricing structures in a connectivity land-grab knowing that, once deployed, the likelihood of displacement is low.
This does not help the cause for satellite IoT. So how does the satcom industry participate without cannibalising its traditional M2M business?
In the connectivity supply chain for IoT, an operator's average revenue per 'thing' (ARPT) must be low-enough to support a decent return to the end-user.
The end user of an IoT system might be prepared to bear operating costs around half the expected value it creates.
The value of IoT sensor data is, as previously noted, totally dependent upon the application, for example:
- Sensor data from a vehicle-tracking IoT application might save fuel costs of around $50 per month.
- Sensor data from a small area of farmland using agricultural IoT applications might achieve combined cost-savings and yield improvements around $500 per month
Assuming in both cases that operating costs are limited to half of the value created, they might sustain an operator ARPT of $25 and $250 pcm respectively. Interestingly, the vehicle-tracker might push more data traffic onto the network, despite returning only a tenth of the value. This illustrates that 'per MB' monetisation strategies need to accommodate as much as two orders of magnitude in price variation to optimise different supply chains. A value-based approach is required to prevent the various IoT markets from cannibalising one another.
GET INVOLVED WITH THE CW JOURNAL & OTHER CW ACTIVITIES
One practical implementation of a value-based IoT pricing strategy is the development of niche supply-chains, whereby operator-approved devices incur a provisioning fee when activated on the network. Within such an arrangement, there might be several classes of 'thing', with each class attracting commensurate provisioning revenues from the value chain.
Device-centric revenues scale-up with a growing population of connected 'things', rather than network traffic. Under this model, it is even possible to imagine systems being sold to the end user with a pre-paid, 10-year connectivity contract included. A payment would then be released from the device vendor to the operator at the point of provisioning.
This works well in a B2B context but implies the need for strong vertical partnerships between the operator, device manufacturer and systems integrator. Such partnerships afford all participants the opportunity to extract sustainable commensurate value from a diverse set of IoT opportunities.
We must consider the breakeven requirements for a new IoT system deployment separately.
If an IoT device has a fully-costed materials and installation cost of $A and sustained value creation of $B per month, its value chain must operate (A/B) months before reaching breakeven. If, just for the sake of argument, breakeven is required in year three to trigger an IoT investment decision, perhaps an upper cost-target for 'things' would be around 30 times the monthly value-creating potential.
Applying this assumption to our previous examples implies:
- A vehicle-tracking IoT supply chain might require the fully-installed cost of a satellite tracker device to remain below $1,500.
- A small agricultural IoT supply chain might require the fully-installed cost of ground sensors plus backhaul to remain below $15,000.
In some IoT cases, there could be a motivation for the supply chain to subsidise terminal hardware if it were proven to stimulate a higher level of service adoption by accelerating the end-user's breakeven point.
We have already advocated the formation of strong partnerships between operators and their device vendors and systems integrators. Terminal subsidies would be transparent to end users and simple to implement (for example, by reducing or deferring a device provisioning fee). In fact, IoT supply chains may find it hard to unpick the capex and opex challenges, unless they are free to behave in this way.
The Alphasat is a large telecommunications satellite primarily designed to expand Inmarsat's existing global mobile telecommunication network. It was engineered and built by Astrium through a public–private partnership (PPP) between ESA and Inmarsat. Using a prototype L-band Advanced Communications Element (LACE) terminal. The experimental terminal established a record-setting 10Mbps return link connectivity from a small L-band antenna.
Generally, it pays to develop IoT opportunities delivering the highest rates of user value-creation per capex dollar spent on the ground equipment.
This is where problems start to arise for satellite IoT as satellite modems tend to be rather expensive. This is partly due to reduced economies of scale compared with cellular, but satcom laws of physics are not particularly helpful either.
A very low-cost satellite-connected 'thing' might cost $80 versus $30 for its cellular counterpart and that yields two possible outcomes:
- On a like-for-like basis, the cellular connectivity platform delivers a faster break-even.
- The total addressable market for satellite-to-node becomes constrained to 'things' able to sustain a minimum value creation of, say, $2 per month or higher (giving an operator ARPT of $1 using the 50 per cent assumption). On the other hand, cellular platforms might support a business case with operator ARPT in the tens of cents bracket.
This has little to do with the relative cost to deliver data from space and everything to do with availability of low-cost satellite terminals on the ground.
For some IoT applications, the characteristics of satellite connectivity might serve to enhance overall value, think: global connectivity - especially in remote areas with no conventional infrastructure, service longevity, higher-reliability, higher-security over a single private network, and this can be sufficient to tip the balance in satcom's favour. If a supply chain cannot extract extra value from those characteristics, the satellite technology platform will always be placed at a disadvantage. But there might be a way to avoid this trap.
Disparity between the value of connecting a 'thing' and the volume of data it uses, means 'per MB' pricing cannot address IoT
Arguably, a more interesting opportunity for satcom IoT leverages terrestrial wireless technology on the ground (Wi-Fi, BLE, LPWAN), with satellite connectivity at the edge of a private network. Within this supply-chain, the cost of satellite ground terminals becomes amortised across tens, hundreds or even thousands of value-creating sensors. Aggregating the total value of 200 sensors has the effect of diluting the satcom capex penalty by a staggering 99.5 per cent. This levels the playing field vis a vis mobile network operators, but only if the operator manages to establish a sustainable value-based pricing strategy first. As we have already discussed, backhauling such IoT deployments with a conventional per MB billing arrangement will not support an optimum supply chain.
The twin pillars of value-based pricing and Satellite at the Edge have strong potential to expand the total addressable market for satellite IoT, in a way that competes with terrestrial connectivity platforms.
Another way to support higher levels of value-creation per terminal is through the application of sensor analytics and big data. Operators might not seem best-placed to directly participate in this opportunity but it is certainly possible for Satellite at the Edge terminals to play a strong supporting role in edge-to-cloud analytics.
We can envisage new ways for the satellite industry to fully capture value from the IoT revolution but seeing how the required changes will be catalysed is less obvious. If vertical partnerships must be formed in each target market, what will be the primary impetus that helps them to form?
Today, the leading satellite IoT companies have aggregate subscriber volumes running in the low millions, with healthy levels of year-on-year growth. The remaining incumbents may soon feel emboldened to participate, faced with a declining revenue per MB and relentless pressure to build capacity.
Perhaps one of the new LEO operators will be quick to establish an attractive business model and scale for satellite IoT, prompting other players to follow, partner or acquire?
It is also possible the terrestrial network operators will take a lead – presenting satellite operators with an attractive opportunity to slot their connectivity platforms into a value-based collaboration within specific private enterprise opportunities, such as oil and gas, or maritime.
Either way, IoT is still very much in its infancy. If we really do need a world with 50 billion connected devices (and most of these could deliver a sustainable value creation), the satellite operators may carve out a strong position to unlock this value. But only if they are free to take a value-based approach to monetising IoT.
After all, people might be happy to buy more megabytes but 'things' can only ever buy the business case.
GET CW JOURNAL ARTICLES STRAIGHT TO YOUR INBOX Subscribe now
Adrian Hillier is a Technology Consultant with more than 20 years of experience leading cutting-edge developments in the mobile communications sector, spanning cellular, LPWAN and satellite systems. At TTP, he assists the Wireless Communications Group in developing business opportunities with clients in the satellite IoT and smart antenna arenas. Adrian previously spent a significant part of his career in the consumer semiconductor industry, developing disruptive modem IP for some of the biggest players in the cellular industry.