We’ve always had difficulty with the term “Internet of Things” (IoT). We believe it is an in-apt phrase, a misnomer for a massive amount of developments that stem from what was commonly called Machine-to-Machine (M2M), and which today is largely Machine-to-Cloud (M2C).
We think, in particular, that likening this set of developments to the Internet is confusing, and may cause the IT and related industries a lot of wasted effort, regarding issues such as standards, security, etc.
IoT – What Does It Mean, To Whom?
Pinning people down on what they mean by IoT illustrates the confusion. In the course of preparing this article we reviewed four recent reports on “The Internet of Things.” A report from Verizon (commissioned from ABI Research) defines IoT as follows: “The Internet of Things (IoT) refers to machine-to-machine (M2M) technology enabled by secure network connectivity and cloud infrastructure, to reliably transform data into useful information for people, businesses, and institutions.”
The first thing to note is that VZ has to go back to the earlier terminology of “M2M” to give any meaning to what it is describing. It seemingly adds two elements, “network” and “cloud.” However, it was always clear the M2M meant that there had to be “network,” so nothing new there. And adding the cloud simply underlines our point that today it is largely M2C.
The fact that IoT is a vague and confusing term has occurred to others as well. A survey published by Telecoms.com Intelligence described IoT as “a very loosely-defined term” largely “shrouded in mystery.”
Also a note of history: The phrase “Internet of Things” is actually attributed to an individual, scientist Kevin Ashton, who coined it in 1999. Interviewed by Smithsonian.com Magazine in 2015, the following exchange occurred:
Q. “You coined the term “the Internet of Things” in 1999, but it can still seem an odd concept. How would you describe it?”
Ashton. “In the twentieth century, computers were brains without senses—they only knew what we told them. That was a huge limitation: there is many billion times more information in the world than people could possibly type in through a keyboard or scan with a barcode. In the twenty-first century, because of the Internet of Things, computers can sense things for themselves. It’s only been a few years, but we already take networked sensors for granted. One example is GPS-based location sensing.”
This answer is not well worded, since rather than defining what he meant by the term he says that we’ve added sensors to computers “because of the Internet of Things.” This is circular reasoning. Furthermore it’s inaccurate, because in areas such as telematics, there already was a good deal of sensor development going on for years in the 20th century.
The Problem Is The Word “Internet”
Now we don’t question that there is a massive amount of development occurring and which will flourish in the future that involves M2M and M2C developments.
We, however, have trouble working the Internet into all of this. The key and most fascinating aspect of the Internet is that it allows everyone’s computer to communicate with everyone else’s computer! (Definition from Dictionary.com “Internet. The global communication network that allows almost all computers worldwide to connect and exchange information.”) This is certainly not the purpose of M2M or M2C applications.
The so-called IoT is in fact a loose number of applications areas, which all have opportunities for beneficially using M2M/M2C technology. Only in structured situations that fit these verticals is it beneficial to have devices communicating to other devices.
McKinsey/GSA Study Illustrates Different Areas
This is illustrated in a study aimed at the semiconductor industry, prepared by McKinsey and the Global Semiconductor Alliance (GSA). While it still embraces the umbrella of IoT, the study delves into six different areas, namely, 1. Industrial, 2. Medical, 3. Smart Cities, 4. Connected Cars, 5. Smart Homes, 6. Wearables.
The McKinsey/GSA study shows that the development cycle and issues to be confronted by semiconductor companies in each of these areas are very different. This is generally a helpful analysis. It makes key points such as: “The readiness of IoT for reaching the growth inflection point varies by vertical.” It also lists how key limitations on IoT development, such as security and privacy issues and low customer demand, vary by vertical.
Findings From Other Studies About IoT Business Opportunity
In another study, Telecoms.com Intelligence surveyed about 1000 respondents with a wide range of connections to the telecom industry. It found inconsistencies in responses, including the fact that the respondents associated the concept of IoT primarily with wearables and with smart homes and connected cars, in other words with consumer markets. However, a majority of the respondents stated that they expected Industrial to be the most lucrative of all the areas of IoT, leading the survey team to conclude: “Curiously, this appears to contradict the conception of IoT being a primarily consumer-focused tech by the responses to the previous question.”
This survey also found that IoT was expected to have little impact on near-term earnings.
A survey of 235 executives in a wide range of verticals by analytics solutions provider Vitria found that: “the top anticipated benefit from real-time Big Data analytics for IoT initiatives is predictive maintenance.” The study concluded that “organizations are initially looking into operational efficiency initiatives, such as asset utilization and cost reduction, as the first wave of project commercialization followed by business initiatives, such as revenue generation and customer satisfaction, as the second wave.”
The aforementioned VZ study forecast 5.4 billion connections worldwide by 2020, up from 1.2 billion devices today. The study presents interesting information about applications particularly in energy and utilities, manufacturing, public sector, and transportation.
However, it compounds the confusion by containing a section entitled, “The Internet of People.” This includes fitness wearables and healthcare devices. The implication is that the Internet of People is simply part of the Internet of Things.
It also raises the question of whether IoT includes smartphones. A good deal of the sensor functionality of wearables can be accomplished by smartphones. The response from IoT coiner, Ashton, quoted above suggests that it does, since he seems to include any computer that contains a sensor.
This last point is related to the underlying, and perhaps purposely unmentioned by these, assumption that scale is related to economic value. The ASP of a smart phone is 200 plus dollars while the ASP of the average “thing” is likely to be in the low two digits if that much. Therefore 50 billion may have scale but the value question is quite open.
Our conclusions are: 1) The terminology of IoT is so widely established that it will not go away, even if it is an inaccurate term. 2) The phenomenon being described is actually a vast opportunity to develop M2M/M2C applications, which vary widely by vertical, business model or other category as to their uses, requirements and commercial and economic potential. 3) We will continue to follow closely the evolution of these applications, particularly those that involve mobile and cloud capabilities.
Links to the studies discussed:
Source of Kevin Ashton quote: Smithsonian Magazine