Massive IoT aims at connecting billion of devices around us (even those deep indoors) that send small amounts of data volumes (not very often) and that work with low power required for long battery life (meaning years). A good example would be connecting smoke detectors in our homes (linked to a home insurance service). Massive IoT is the Revolution, meaning IoT being deployed in most sectors on a massive scale (instead of applying IoT in specific projects or segments).
Applications of massive IoT include diverse use cases as logistics and tracking (asset tracking), utilities (smart metering), smart building (smoke alarms), agriculture (climate-field monitoring, livestock tracking), industrial (process monitoring), retail (store operations), smart cities (bicycles, parking) and connected wearables.
There are several connectivity technologies that are enabling massive IoT, some covering the wide range like either cellular-IoT (NB-IoT, LTE-M, 5G) or unlicensed LPWA (Sigfox, LoRa) and some other for short-range as mesh networks. However, there is no one-size -fits-all solution, and the most likely choice would be a hybrid one considering complementary connectivity technologies to best serve the requirements of the specific use case.
Massive IoT, with 400 million wide-area IoT connections at the end of 2019 is expected to become 50% of wide-area IoT connections by 2025, i.e. 2 to 2.5 billion connections depending on the source. (Berg Insight) (Ericsson)
This business is trying to take off but not yet living up to the expectations, except for China that is leading the global adoption of massive IoT. One of the main reasons is not understanding that massive IoT is a different kind of business, with high volume and lower ARPU (average revenue per user), that requires a different approach to machine-to-machine (m2m) from main IoT operators as Telcos.
Forecasting boosterism on IoT connections
Now it’s early 2020 and 20 billion connected devices are surrounding us…aren’t they? (sorry I can’t see them).
Projections on the number of IoT connected devices from analysts and companies have had significant variability among sources and time. The first big figure I could remember is the 50 billion connected devices by 2020 predicted by Ericsson and Cisco in 2010 and 2011. Today we know that despite being unrealistic, that figure created a great expectation about the potential of IoT business and it has contributed to the development of IoT, as most of investors and companies bought it in the past.
IoT connection forecasts are biased towards the supply and despite that the actual figures are finally lower than forecasted, the growth rate becomes a good indicator for potential opportunities.
There were around 11 billion IoT connections by the end of 2019: over 1 billion for wide-area IoT, including 400 million for massive IoT. An additional 14 billion connections have been predicted by 2025 (comms department have a new marketing claim to play with, “25 billions for 2025!”).
Fig 1. IoT connection (billion). Source: Ericsson Mobility report 2019
Non-cellular IoT technologies, Sigfox and LoRa hold the leading share of LPWA networks connections at least through 2022 over cellular technologies (note that figure 2 only focus on cellular IoT and it doesn’t consider either non-cellular or short-range technologies). These unlicensed LPWA connections are forecasted to reach 300 million by 2023. (Berg Insight)
Fig 2. IoT connections by segment and technology (billion). Source: Ericsson mobility report 2019.
Massive IoT: Connecting billions of devices, small amounts of data volumes, (mostly) sent infrequently, low power required for long battery life (years not days, weeks or months).
Broadband IoT: will need high throughput and/or low latency.; large data volumes.
Critical IoT: will require ultra-high reliability/availability and very low latency. Industrial automation (and robotic surgery) will require time sensitive information delivery and precise positioning of devices.
One of the technologies better suited for addressing the density challenge of massive IoT is 5G, that aims to connect millions of devices per square kilometre. Unfortunately, because of its roadmap availability will have a limited short-term impact on IoT, but will enable great opportunities in massive IoT (when mMTC, massive machine time communication, is available), broadband IoT and Critical IoT.
Revenue growth slower than connections
The global cellular IoT communications market is growing at a steady pace, and the connections are growing quickly (but lower that forecasted). There is a considerable IoT revenue growth, but IoT revenues continue to grow slower than connections. Besides, IoT revenues are still a small part of operator´s business, as IoT accounts for around 1 percent of total revenues for most operators, except for Vodafone, AT&T and China Mobile.
Global IoT revenues increased by around 25 percent during 2018, while the monthly ARPU dropped by a similar order. Excluding China, the trend was less dramatic with revenues growing by 16 percent and ARPU declining between 5-10% per year. (Berg insight)
A very interesting insight is that there is a negative correlation between growth in connections and monthly ARPU. IoT operators that achieved an increase in the number of connections should expect a reduction in the monthly ARPU, and one of the reasons is that the greater part of net additions is the cost of IoT devices. At this point operators would need to be more creative and develop better services on top of IoT connectivity to balance the ARPU.
Drivers to take off
Some years ago (2016) I made some detailed analysis and presentations regarding the demand for massive IoT use cases, that as you may know are enabled by LPWA (low power wide area networks). One of the main conclusions was that LPWA demand was not yet living up to the expectations and that the market needed to focus on getting lower e2e service price by i) reducing the connectivity module cost ii) getting lower device cost iii) building-up a proper scalability and iv) sorting out the lack of interoperability among all the elements of the IoT ecosystem.
Now, 3 years later, the connectivity module cost has been sorted out ($2 to $3 for NB-IoT modules and under $1 for Sigfox modules). Interoperability issues are work in progress but getting results that are quite promising. By now, scalability requires devices connected reliably and safety and it depends on many factors as: a) knowledge of business processes and a proper business case defined for each use case b) most of the scenarios requires complementary connectivity to success c) high-end security is a must to guarantee the end to end service.
But the real issue here is that there are no low-cost IoT devices at the proper scale (although there are some exceptions) to support the take-off of massive IoT business. And it is not only the cost of the hardware equipment (as sensors) that compound the device that cause this issue. Most of the device manufacturers that develop great IoT devices but keep supplying the m2m market, are afraid of being affected by cannibalisation (the premium m2m device would be cannibalised by a growth in sales of its cheap IoT device).
A good example is the tracking use case (asset, cattle, people). A GPS-device tracker cost around a few hundred dollars for traditional m2m business, e.g international freight container. But if you need a tracker for a massive IoT use case e.g. either tracking cattle or pallet, the same vendor charges you a similar amount (hundreds of dollars) despite the business case requires a device tracker costing a few dozen dollars. This issue could be addressed in two main approaches i) device vendors reducing the cost of IoT devices when targeting massive IoT use cases, betting on a high-volume business ii) innovation on the IoT device market. A good example of innovation is Sigfox Atlas, a location service using network location (Native), and WiFi or beacons devices for better location accuracy.
A different kind of business, not machine-to-machine
Over the last years, m2m has mostly been focused on connecting a few high value devices and gathering data to be displayed somewhere (e.g. vending machines sending out inventory information). IoT goes beyond m2m (including it) endowing devices with the capability to talk to each other creating new applications and transforming businesses (according to the GSMA definition: The Internet of Things describes the coordination of multiple machines, devices and appliances connected to the Internet through multiple networks). Massive IoT is the core of IoT deploying its application on a massive scale.
Main IoT operators as Telcos have IoT units that have evolved from traditional m2m business (fleet management, parking meters and managed connectivity). Despite some of them have created separated business units for the IoT/m2m, the issue remains because they are still using the m2m ARPU as the main reference to price an IoT service and applying the machine-to-machine philosophy to operate a different kind of business. It is not just the device vendors that have a cannibalisation issue when dealing with both m2m and massive IoT use cases, IoT operators do as well. It is understandable that it is difficult for them to make a bold move betting on a lower ARPU business (despite of its high-volume potential), but it is the way to make this business taking off.
A bold example is Objenious from Bouygues telecom in France. In 2016 Bouygues unveiled a dedicated IoT subsidiary focused on massive IoT business (LPWA) and aimed at developing solutions covering every segment of the IoT value chain, including devices and specialised teams (even sales) to be more effective. They have developed a catalogue of B2B and B2B2C services to address a number of massive IoT services. In September 2019 Bouygues announced the reunification of the m2m business and LPWA business under the Objenious umbrella. Luckily, they learnt a thing or two from the freedom to do things in a different way to the core business. It is a good example to keep an eye on it.
Therefore, it is key to have a proper organisational structure that allow a befitting space for both m2m and massive IoT business, with proper coordination between them to avoid cannibalisation and providing a clear differentiated offering to the market and clients. This organisation would need specialised teams for each business, understanding both the business processes and specifics of each application. Working this way these teams will be capable to develop successful end-to-end use cases and IoT applications with a clear value proposition and a viable business model.
The IoT short-range opportunity
The short-range segment consists of devices connected by unlicensed radio with a typical range of up to around 100 meters, such as Wi-Fi, Bluetooth and ZigBee. This category also includes devices connected over fixed line local area connections.
By 2025, 80% of IoT connections (nearly 20 billion) will be supported by short-range IoT (from 90% in 2019). (Ericsson)
Rather than competing with each other, these technologies could complement the offering of wide-area IoT use cases in local environments, covering the “last 100 meters” connectivity. It is a win-win scenario with wide-area connectivity providing a backbone for mesh network to reach farther and at the same time short-range technologies creating more demand for cellular and non-cellular connectivity operators.
Actually, there are already some examples of hybrid connectivity use cases in asset tracking or smart metering, using LPWA plus WiFi or Bluetooth. Besides, I find really interesting the cases using mesh technologies (as ZigBee) and their application to smart home. Zigbee alliance is developing several initiatives for smart and connected home with members as Amazon, Apple, Google and Ikea, highlighting the development of new open standard for smart home device connectivity. Surprisingly there are no many Telcos in that list (actually only Comcast) and it seems to me a good missed opportunity for Telcos to better engage in this smart home business, generating more demand and revenues.
IoT powered by Artificial Intelligence (AI)
IoT does not make sense without data and insights. IoT is all about data, to create value from connected devices and improve business outcomes. IoT devices provide data sets that through big data, Artificial Intelligence and machine/deep learning become solutions to business problems and create new opportunities.
IoT Big Data-AI is a key piece to address the market opportunity “beyond connectivity” (security, blockchain, cloud and edge computing are part as well despite not being included in this analysis at this point). The growth in addressable revenue opportunity by going beyond IoT connectivity is 5 to 15 times the IoT connectivity. (GSMA Intelligence and PwC analysis)
In 2025, the global addressable market for Mobile Network Operators (MNOs) in IoT Big Data beyond connectivity is estimated at approximately $386 billion: 30% in professional services, 29% in analytics and big data software, 20% in applications, 13% in Cloud and hardware, and 8% in platforms. (GSMA Intelligence and PwC analysis)
Some MNOs are beginning to apply their IoT, Big Data and AI capabilities to provide different use cases across industries as transport, logistics, retail, agriculture and smart cities but it’s just the starting point and they need to become global services to make a real impact.
Some suggestions for Telcos to develop massive IoT and achieve a steady growth in IoT
IoT and particularly massive IoT, could become the most valuable market in the industry for a decade and therefore there are so many players claiming their slice of the cake. Telcos were in the best position to become the IoT main players transforming themselves from dump pipe operators, building a new business on their core capability, a reliable and secure connectivity (and growing beyond that). Although there are a few examples of Telcos doing things right, most of them are missing the boat again just anchored in their old m2m business. Telcos would need to do additional things to grow IoT business by the target, going beyond connectivity, reinventing themselves and creating an ecosystem of partners to capture better opportunities. I suggest here five main complementary approaches and opportunities for them:
- Grow the traditional machine-to-machine business focused on low volume and high value segments by acquisitions and partnerships in one or two new businesses. This complements the traditional organic growth of m2m in segments as fleet management or international freight
- Bet on massive IoT business, addressing and developing it properly. This considers solving some of the issues discussed above:
- Focus on successful use cases in the market for massive IoT, as smart metering or asset tracking
- Provide hybrid IoT connectivity offering (cellular IoT + non-cellular + short-range) to implement the short-term massive IoT offering, providing the best solution adapted to each scenario
- Invest in cellular IoT network deployment: either NB-IoT deployment in more countries or accelerate 5G mMTC (in principle not available in short term). Take into account important topics as the impact of both the introduction of eSIM/iSIM and the merging of NB-IoT and LTE-M into 5G standards
- Implement a proper organisational structure, even exploring the option to create a separate unit for massive IoT
- Lower device pricing (eg. trackers), as one of the main contributor to growth and acceleration. Partnership, acquisitions and internal development are complementary alternatives to explore
- Develop new product and businesses through open innovation
- Invest in high-end security to protect privacy and integrity of IoT users
- Reinforce the beyond connectivity strategy with new products on IoT powered by Artificial Intelligence
- Create products with a common vision and strategy (IoT, big data, AI). This is an organisational challenge
- Enrich the data from other sources through partnerships, to be combined with device data and get better insights
- Create AI powered products with predictive reasoning (the ability to forecast events with a probability that increases over time) and using machine learning and deep learning technologies
- Leverage/Create partnership with China companies to learn from IoT Big Data-AI use cases. China is leading the global adoption of massive IoT and have become leaders in applied AI
- Explore the benefits and applications of blockchain and edge computing applied to IoT
- Explore the short-range IoT opportunity in selected markets
- Analyse the benefits to use these technologies for use cases as remote control, security, home automation, energy management, positioning and automotive
- Prepare a business case to evaluate the technology/ies and use cases most likely to success. This opportunity seems the most radical approach, so some further analysis needs to be prepared before testing
- Explore joint use cases with WiFi operators, as a combination of connectivity options (LPWA could be used for connecting the devices and WiFi for processing the data, as a wireless backbone or for devices with less demand on low power consumption)
IoT figures and forecast from:
Ericsson. Mobility report 2019
Berg Insight. The Global M2M/IoT Communications Market. IoT Research Series 2019
ABI research. Low-Power Wide Area Network Market Data 2019
GSMA.The IoT Big Data Revenue Opportunity for Mobile Operators. October 2019
