How big is the Internet of Things? (Part 2)
This two-part article discusses the value of the Internet of Things. We take a non-crystal ball approach to measure IoT’s tangible and intangible value, and make an estimate of its long term value based on real market data.
In the first post, we explored why the IoT needs to make the shift from a technology to a value driven business to meet its growth expectations.
In the second part, we discover how IoT is evolving towards zero marginal cost and infinite value, and why a standards-based horizontal IoT network and platform, on which a broad variety of devices and applications share functions and data, will fuel its exponential growth.
Towards zero marginal cost and infinite value
In “The Zero Marginal Cost Society: The Internet of Things, the Collaborative Commons, and the Eclipse of Capitalism,”American social-economic theorist and activist Jeremy Rifkin describes how new technologies, such as 3D printing, green energy, and the Internet of Things, are quickly moving us to an era of nearly free goods and services (and, according to the author, the eclipse of capitalism.)
“Billions of sensors are being attached to natural resources, production lines, the electricity grid, logistics networks, recycling flows, and implanted in homes, offices, stores, vehicles, and even human beings, feeding Big Data into an IoT global neural network. Prosumers can connect to the network and use Big Data, analytics, and algorithms to accelerate efficiency, dramatically increase productivity, and lower the marginal cost of producing and sharing a wide range of products and services to near zero, just like they now do with information goods.”
The programmable world will have a immense impact on many industries. Take for example the logistics and transport sector: access to real-time data, like the location of goods, current weather conditions, traffic flows, and up-to-the-moment information on warehouse inventories and capacities, will dramatically reduce the marginal cost of production, storage and delivery of goods.
At the same time, the convenience brought by IoT technologies will have a tremendous added value for a lot of companies and consumers, since it will boost productivity, increase efficiency, and improve the end-customer experience. As such, it would be a fair statement to make when saying that the long term value of IoT is nearly infinite.
In the next section, we will, based on real market data, make an estimate of this long term value.
Calculating the long term value of IoT
Let’s begin the quantitative analysis phase of this paper by looking at Moore’s law. Moore’s law is typically calculated as the number of components per integrated function in an electronic circuit. The theory states that the number of components per given functionality in an integrated circuit doubles every two years. This translates in a number of ways – circuits can get faster at a fixed size, smaller at a fixed speed, lower cost at a fixed speed and size, or infinite combinations of any of the three factors. The end product is optimized based on its intended use.
We begin to build our Moore’s law curve by starting in 1971 with the Intel 4004, which had 2,300 transistors per CPU. Then we extrapolate that forward to 2021. On the chart that follows, we show the density and cost evolution between 2013-2020 in billions of components per function.
Figure 8: Moore’s law for the IoT
Between 2015-2021, the performance of electronic circuits will increase by 8 times. This means that the cost of a fixed function will decrease by the same amount (8 times). Sensors will become much less expensive and smaller, and can be widely deployed at low cost. The data collected by these sensors can be transmitted more cost effectively, and analytics computing power applied more quickly and cost effectively.
Next, let’s look at quantifying Metcalfe’s law. The objective is to be able to compare Internet users and their implied value to IoT devices to be able to make a relative comparison of potential. We begin with the number of Internet users from 2000-2016. We also obtain a projection of Internet users for 2020. In addition, we plot the number of connected “things” and their respective values squared. In this example, conservative numbers for connected “things” are used, reaching 21 billion by 2020. Some projections go as high as 100 billion “things”. The latter value exaggerates the value of IoT beyond what is used in this example. We plot these values on the following chart, and also plot their “value” based on Metcalfe’s theory with 2013 as the baseline.
Figure 9: Metcalfe’s law for IoT
Based upon this extrapollation, we conclude that by 2020 IoT will have a value, based on the number of connections, that is 7.5 times that of the Internet in 2020 and about 36 times the value of today’s Internet, assuming the node values are equal.
As the number of devices connected through IoT increases due to Moore’s law, and subsequently drives down the cost of the technology required, the value of IoT accelerates. With costs dropping and value increasing exponentially, new business models become viable and sustainable. This is when IoT moves from being a technical “wow” to an actual business, and in other words, it “crosses the chasm”. And acceleration continues because now it can be used to reduce transaction costs and push ineffiencies out of the system.
A standards-based horizontal IoT platform fuels exponential growth
Many of the early IoT applications came to market following a vertical model, in which each solution was implemented on separate infrastructure, which needed a new and/or tailored IT development effort, and used a separate management and execution platform. Such a siloed approach not only has an impact on development, deployment and operating costs, but it also constrains the opportunity to realise meaningful economies of scale, develop solutions that can apply to more than one vertical market, and benefit from application interworking and data sharing.
Although, at first sight, IoT is all about vertical solutions that beg for a vertical model, there is a clear business rationale for an “any network, any device, any application” platform with a common set of service capabilities, standardized interfaces, and open APIs enabling faster growth and innovation, as a broad variety of devices and applications can use common infrastructure, and share platform functions and data.
Actually, a horizontal IoT platform will boost the network’s value as defined by Metcalfe’s law. If we take the example configuration below. When N different vertical applications (Ai) are deployed on N different “stovepipe” platforms (Pi) with M connected devices (Dij)each, then the discrete y-value of Metcalfe’s curve is equal to N*(M+1)2.
Figure 10: Metcalfe’s law for N vertical applications
But, if we assume a common, horizontal platform that connects all N applications to all N*M devices, then the network value explodes exponentially to (N*(M+1))2.
Figure 11: Metcalfe’s law for N applications on a horizontal platform
Furthermore, a horizontal platform with open interfaces and standardized protocols will allow IoT service providers to reduce investments, lower operational expenses, scale faster, speed up time-to-market, facilitate partnerships, and deliver a better customer experience. While device vendors and application developers can concentrate on their real differentiators, rather than on re-implementing the common functions that will be provided by the platform.
Moore’s and Metcalfe’s laws drive the chasm crossing and in doing so, enable increasing value while catalyzing the continuing drive toward zero marginal cost and into a dynamic world that continually adjusts and optimizes seemingly on its own. This world comes about when sensing, connecting, and computing costs no longer prohibit ubiquitous deployment and when horizontal IoT infrastructure, like Nokia’s IMPACT platform, help maximize synergies between an ever growing number of devices, data records and applications.
By understanding in qualitative and quantitative terms how the confluence of all of these factors drives IoT, it’s quite easy to see why many predict the value of the Internet of Things to be in the trillions.