Home > Progression from M2M to the Internet of Things: an introductory blog

Progression from M2M to the Internet of Things: an introductory blog

Welcome to Analysys Mason’s monthly blog post on trends and happenings in the Internet of Things (IoT) world. When Bosch asked me to write this blog series, I was excited. I have known Bosch in my home and my car, and now I get to have my name on the Bosch website. I am going to try to convince Bosch to name a new high-powered, electric hand drill in honor of me, but I do not know if I will be successful.

Over the next 12 months I am going to write a series of blog posts and respond to your comments about all sorts of IoT topics. I will discuss various industry sectors’ adoption of connected devices and applications; look at new business models for IoT; ponder some M&A activity; talk about platforms; and highlight some solutions adopted in various regions of the world. Certainly let me know if you have ideas or thoughts for blogs as well.

Machine-to-machine (M2M) is a technology that uses a device attached to a machine to capture an event which is relayed through a mobile phone or fixed line network to an application that translates the event into meaningful information.

The Internet of Things (IoT) is the next generation of the Internet based on the Internet Protocol (IP). In addition to today’s Internet it facilitates via end-point devices the automatized use of data generated by sensors or devices and access to actors. IoT solutions are deployed in sectors including automotive, transportation, smart homes, energy, utility, security, surveillance, public safety, financial services, retail, healthcare, industrial, warehousing, and distribution. Sometimes IoT is used synonymously with the term machine-to-machine (M2M). Undoubtedly there is some overlap, but I will attempt in this blog to show the progression from an M2M to an IoT world.

How large is the market for connected things?

The number of device connections worldwide will increase from 100.4 million in 2011 to 2.1 billion by 2021, growing at a compound annual growth rate (CAGR) of 36% during the forecast period (for more information see Analysys Mason’s 2011–2021 M2M worldwide forecast report). We forecast M2M rather than IoT, because generally we are still in an M2M world. But it still gives us an indication of the magnitude of the industry.

In 2011, 69% of these devices were in developed regions of the world although we expect this percentage to decline to 59% by 2021 as enterprises in emerging regions start to notice increasing ROIs from adopting M2M solutions. In 2011, 80% of M2M connections were over mobile networks – primarily 2G and 3G –and this percentage will increase to 93% by 2021, because the ongoing costs associated with M2M over mobile networks are generally less expensive than fixed networks (see Figure 1).

M2M device connections, developed markets’ share of connections and fixed networks’ share of connections, worldwide, 2011–2021

Figure 1: M2M device connections, developed markets’ share of connections and fixed networks’ share of connections, worldwide, 2011–2021 [Source: Analysys Mason, 2012]

So we know this is going to be a large market opportunity. In fact, some of the largest industry sectors for IoT include the automotive/transport sector; various applications associated with the smart home including security; the industrial sector; and the utility/energy sector.

Now let’s talk about this movement from M2M to IoT. There are three layers of the M2M/IoT supply chain: hardware, connectivity and applications. See figure 2.

M2M/IoT supply chain [Source: Analysys Mason, 2012]

Figure 2: M2M/IoT supply chain [Source: Analysys Mason, 2012]

As we move from an M2M world to IoT, each layer of the supply chain experiences a metamorphosis.

Hardware – M2M hardware has historically been specialized, expensive and possessing low processing power. There have been many specialty vendors creating a relatively small amount of this equipment. In an IoT world, we expect hardware will become more powerful with increased processing capabilities and intelligence. We also anticipate a heightened emphasis on device power management and an overall consolidation of the market.

Connectivity – M2M connectivity has historically been provided over a combination of fixed and mobile networks with pricing that has been undifferentiated from traditional residential or enterprise connectivity service. Levels of quality have varied greatly. In an IoT world, we expect solutions to take advantage of mobile ubiquity and increased network quality and speed. We also anticipate an increased reliance on global tariffing to help facilitate the roll-out of international solutions.

Applications – M2M applications have historically been customized, services-heavy deployments. Enterprises have had minimal data analytics capabilities and have sub-optimally been using the data flowing off M2M devices. In an IoT world, we expect cloud applications enabled by virtualization to make application deployment across common platforms feasible. We anticipate better data aggregation and analytic tools to drive increased cost savings and innovation for enterprises adopting IoT solutions.

Each of the three layers of the M2M/IoT supply chain will transform and this will enable a series of changes in the industry over the next 5-7 years. Let me provide a few.

  1. First, we are going to see a proliferation of new applications. These applications will change the way we track, monitor, evaluate, protect and improve the things in our lives. Applications might include home energy management, predictive maintenance, surveillance, online interactive laboratories, intelligent vending machines and interactive advertising.
  2. Second, we will interact with our automobiles, other vehicles and our homes in pioneering ways. Machine-based intelligence coupled with ever faster processing power and connectivity will make our homes and vehicles into epicentres of applications-rich interactions as the Internet crosscuts potentially all application domains.
  3. Third, IoT will drive businesses to innovate and create revenue-generating ideas. Today, many M2M solutions are focused on reducing an enterprise’s costs. But in the future, we will see embedded connectivity driving more product and service innovation.
  4. Fourth, we are going to see all sort of new partnerships between technology and equipment vendors; communications providers; application vendors; and services companies. IoT is going to encourage businesses to think “outside the box” and change business models. It is going to encourage equipment vendors to offer software and services. It is going to encourage services companies to price their offerings in new ways. It is going to encourage manufacturers to change their supply chain dynamics

What is your guess at some key trends in the IoT industry over the next 5-7 years? Any speculation about some of the new applications we will see in our homes, businesses and cars?

Thanks for reading this first post. I will be back shortly with a post on the market of IoT platforms and systems management.

Article by Steve Hilton

Steve Hilton is the founder and Managing Director of MachNation, an insight services firm covering the future of the Internet of things (IoT). Prior to founding MachNation, Steve has worked as lead analyst for Analysys Mason's Enterprise research program until December 2013. He has also held senior positions at Yankee Group, Lucent Technologies, TDS and Cambridge Strategic Management Group (CSMG). Steve’s primary areas of specialization, which focus on large and small enterprises, include IoT, fixed and mobile communications services, M2M, cloud services, and sales channels. He has 20 years' experience in technology and communications marketing, with an IoT specialization for 7 years. Steve holds a degree in economics from the University of Chicago and a Master's degree in marketing from Northwestern University’s Kellogg School of Management.
more articles by this author

Leave a comment

  1. from Jan Behrmann   /   October 12th, 2012 at 22:14

    Steve,

    Thanks for the good introduction!

    In my oppinion there is one key-statement in your article:

    “[...] IoT will drive businesses to innovate and create revenue-generating ideas. [...] in the future, we will see embedded connectivity driving more product and service innovation.”

    This has to be burned into the minds of the management staff.

    My personal guess is, that the potential of IoT (or M2M) will only be effective with the right amount of creativity and the courage to think uncommon and not to focus only on the next quarter.

    And I believe, that lots of innovative products can be introduced with the technology available today – even if there is still missing interoperability. This will boost IoT, not the enterprose use.

    I do not see a reason to wait.

    Regards
    Jan

  2. from Steve Hilton   /   October 12th, 2012 at 22:35

    Jan, I agree with you. IoT will have innovation/revenue generating impacts for many firms. I’ve seen it already. Take the case of Southwest Airlines — a US-based, national airline. The cargo shipping division of Southwest Airlines is using IoT/M2M to create new track and trace services for its cargo-shipping clients. It’s a great solution from an enterprise with lots of foresight. We wrote a piece of research on Southwest Airlines’ cargo tracking solution at:

    http://www.analysysmason.com/Research/Content/Comments/M2M-Southwest-Airlines-Sep2012-RDME0/?bp=http%3a%2f%2fwww.analysysmason.com%2fSearch%2f%23query%3dsouthwest%2bairlines%26access%3dAll%2520content

    Thanks,
    Steve

  3. from Charalampos Doukas   /   October 13th, 2012 at 22:03

    Steve, thank you for this thought-triggering introduction (and congratulations to Bosch for this initiative)!

    The following years will be very crucial for the commercialisation of IoT and penetration in home, business and other domains. Most likely, we will see home routers support additional wireless technologies, new standards for interoperability between services and devices and probably a dominant wireless technology that will feature/satisfy main IoT networking requirements (low power, low cost, etc.).
    The most successful applications will probably be those that will enable data collection and integration from different resources (e.g., user context, quantified self, web, etc.) and drive smart interactions with users (automation, decision support, etc.).
    Many challenges however to overcome till then!

    Looking forward to your next post!

    Best,
    Charalampos

  4. from Thomas Nicholls   /   October 14th, 2012 at 15:10

    Thanks for a good article Steve (and Bosch).

    The dawn of IoT has been announced many years ago, but no one has really been thinking out of the box to come up with a new type of solution that addresses the specific connectivity requirements. At Sigfox we believe that it’s time for an new type of cellular network, that specifically addresses the IoT market. The network must be cheap, eco-friendly and with a broad coverage for world wide solutions.

    Our network is already in use and is being rolled out world wide. We’re hoping to be the connectivity disruption that we believe this space needs.

    Looking forward to the next article.

    Regards

    Thomas

  5. from Steve Hilton   /   October 15th, 2012 at 15:17

    Charalampos,

    Thanks for the comment. I like your comment about “The most successful applications will probably be those that will enable data collection and integration from different resources (e.g., user context, quantified self, web, etc.) and drive smart interactions with users (automation, decision support, etc.).”

    I agree with you. The mashing-up of information has proven itself a powerful tool for application vendors (including Google). I think our IoT world is poised for this kind of change. But you’re right, first we need to connect all these things in our world!

    You mention that home routers will support different technologies. Any more thoughts on that idea?

    Thanks again!
    Steve

  6. from Steve Hilton   /   October 15th, 2012 at 15:25

    Thomas,

    Thanks for the comment. What applications/industry sectors do you see as feasible for the types of networks envisioned by Sigfox? Sigfox provides a solution over unlicensed spectrum with very low throughput (you call it ultra-narrowband).

    One of the cool thing about IoT is that there are applications applicable for all sorts of network topologies.

    Let us know.

    Thanks,
    Steve

  7. from Mike Mills   /   October 16th, 2012 at 14:45

    Great article Steve! I agree with you and Charalampos about the importance of data in an IoT platform. We’re just starting to make announcements about our new IoT platform, http://www.grovestreams.com. All of our initial designs centered on managing and analyzing large amounts of data quickly. In my opinion, extracting useful information from a tsunami of data from many different sources so that organizations, people and devices can act on it quickly will be one of the primary measurements of a successful IoT application. It’s an exciting time for IoT. I’m glad to see you and Bosch dedicating time to it.

  8. from Steve Hilton   /   October 16th, 2012 at 14:52

    Thanks for the comment, Mike, about Grovestream. Doing something with all those IoT data is absolutely critical. Do you think there is a way to assemble all those data and make meaningful decisions or take appropriate actions WITHOUT having humans involved in the process? One of the tenants of IoT/M2M is that by removing human latency from decision-making we make business processes more cost efficient.

    What do you think?

    Thanks,
    Steve

  9. from Thomas Nicholls   /   October 16th, 2012 at 15:27

    Thanks for the question about Sigfox Steve.
    As you correctly note, Sigfox targets the low throughput applications. We therefore do not aim to cover all IoT/M2M use cases and rather see ourselves as complimentary to high throughput networking solutions.
    If for instance you wish to transmit a video feed, then Sigfox clearly isn’t what you need. If however you wish to transmit data such as temperature, location, humidity, presence, health data, etc., then Sigfox makes sense. Using a low throughput network allows us to cover these use cases with very low power consumption, low cost and independence of fixed networks, which is not possible with more complex networking solutions.
    We also see quite a few actors being attracted to the advantages we provide and actually re-structure their services to be compatible with a low throughput connectivity.
    The industry is evolving very quickly these days and we are more than excited about what the future will bring. New applications and use cases seem to pop up every day!
    Thanks
    Thomas

  10. from Mike Mills   /   October 16th, 2012 at 16:13

    Steve, I believe humans will always be involved one way or another until AI advances enough. That being said, today, a person can model an environment and then many devices can use the model to get real-time results. For example, a human can sit down and design a spreadsheet that calculates a monthly energy bill based on one minute energy consumption for a certain month. The spreadsheet can be run each minute as the energy data comes in and the monthly bill can be monitored in real-time as the data comes in. So it’s easy to see how a device can obtain the monthly bill data each minute and make decisions based on it. Now, it’s another thing to have the device actually design the spreadsheet. So yes, I see humans doing the modeling for years to come and devices and other humans using the results of the model to make real-time decisions. That’s the paradigm we’ve designed into our system – at least until AI advances enough to replace the modeler :)

  11. from Charalampos Doukas   /   October 16th, 2012 at 17:58

    Steve,

    I am glad to see that you (and the rest of the contributors in the comments thread) are on the same track about the IoT vision and the requirements/challenges!
    About home routers: Wifi is the only home established networking technology for connecting devices like computers and smartphones. But as many will agree, it is an overkill for IoT and still expensive for small, low-cost IoT devices (and they have to be low cost to penetrate the market). In addition, it comes with drawbacks like power consumption and range in real world scenarios. Sigfox has a very good vision on this by providing a ubiquitous network technology which can be a good replacement for the expensive and power-hungry GSM technology. But until it is widely deployed, and for the more throughput demanding applications (like the Quantified Self: some biosignals high resolution for good assessment), there has to be an alternative RF technology. I don’t know if it will be XRF, Z-wave, ZigBee or a brand new IoT-driven technology, (or a combination of them) but for it to succeed it has to be integrated into home routers. Right now, as a consumer, if I needed to buy two different IoT devices, one using Zwave and one using XRF, I would really be skeptical about needing two different gateways attached to my home router. But if my home router would support those interfaces be default, I would not bother that much!

    @Mike: Congrats on your new IoT platform! I believe data mining and pattern recognition is also very crucial for making something meaningful out of IoT. During my research on machine learning I have realised that depending on the problem and the data, there can be techniques that can perform really good in decision making and finding patterns. As usually with computer aided systems, and especially for critical applications, such systems can only be considered supportive and assistive in making decisions: I think it is more that enough to be able to interact smartly with users and suggesting actions, not driving them automatically!

    Best,
    Charalampos

    PS: Seing this thread getting very interesting comments and triggering further discussion, maybe guys from Bosch should consider organising something like an open forum or a google group where people can share thoughts about posts and interact better with authors and commenters.

  12. from Steve Hilton   /   October 16th, 2012 at 18:45

    @ Thomas — the industry is definitely changing quickly. There is clearly a need to recognize that all applications don’t need to run on the same network. Sounds like that’s the Sigfox approach — it’s all about segmentation of the IoT/M2M marketplace.

    Is Sigfox similar to the “old” two-way paging networks?

    Thanks again,
    Steve

  13. from Steve Hilton   /   October 16th, 2012 at 18:52

    @ Mike — I think you are correct. I can’t imagine replacing all human interaction completely, but I’d like to think we can minimize interactions if possible with some simple rules-based models. I’m no AI expert, but I have a PhD friend over in the UK who is an AI expert. I should talk with him and determine when we can all retire at 36yo and have computers do all the work for us.

    I like your utilities/energy example. Do you know if utilities that have smart metering programs are actually capturing (and using) data on a by-minute basis to make decisions? I don’t know if there are any utility/energy sector experts reading this chain.

    Thanks again,
    Steve

  14. from Steve Hilton   /   October 16th, 2012 at 19:00

    @ Charalampos – thanks for the posting. 802.11x isn’t built for IoT. I agree with that statement for any number of reasons. I think GSM/GPRS networks have their place in the IoT world, but don’t have to be used exclusively.

    It makes sense to me that an existing home/enterprise router would support 802.11x and other standards. That would make most sense from a consumer perspective. I wonder what the dominant vendors — Cisco et al — think about this issue. There’s a new IoT group at Cisco. I’m trying to set-up a call with them to understand their direction and objectives.

    thanks again,
    Steve

  15. from Mike Mills   /   October 16th, 2012 at 19:37

    @Steve – I designed and built energy billing and forecasting system years ago so I have some knowledge in that area, but I’m sure things have changed a bit since then. I believe smart meters and regular meters can record data down to the minute or less but the problem is handling the amount of transactions and data gathered from millions of meters. Meter companies, such as Itron, now have meters that can broadcast their usage to a local router (not sure of the protocols) so that a home owner can view their usage in near-real time. But, as a home owner, it’s nice to see how much energy I’m currently using, but what I really want to know is how much is it costing me and how much each item in my house is costing me. For this to happen, the meter would need to know how the bill is calculated (and note that the calculation can change over time). Thus a centralized modeling platform is needed for the modeling. The actual calculation could take place locally or in a system like GroveStreams. My opinion is that industries are demanding that the energy industry give them their usage and billing information down to the minute or less but the energy industry just doesn’t have the technology to provide that information from millions of meters to millions of users. That’s one area we think we are addressing. One of our demo apps is collecting one second data from a zigbee smart plug behind a refrigerator, transmitting to a Digi router. The Digi router uploads one second usage every ten seconds and based on how we modeled the smart meter data streams, we display usage down to the second and we also display monthly costs, which are calculated every few seconds. The results are available to users or other devices. Some results are in real-time as the data is uploaded; more complex calculations are done in near-real time. You can view our smart plug sample dashboard here: https://www.grovestreams.com/observationStudio.html?&org=00000000-0000-0000-0000-000000000001&view=dashboard&itemUid=45c1ad1d-b022-35eb-b92d-3d91adce10f3&itemName=Refrigerator%2520-%2520Smart%2520Plug%2520Sensor&guest=true

  16. from Steve Hilton   /   October 17th, 2012 at 14:58

    @ Mike – The smart plug is a good example and I like the ability to capture usage data and link it to energy tariffs to provide (near) real-time calculations of monthly energy spend. That’s a great tool.

    I agree with you that energy companies will find themselves awash with massive quantities of energy usage data and little ability to present it to customers (residential or commercial) in any meaningful way. As you’ve said, the answer exists in understanding IoT customer needs and requirements first, and building a solution to address those needs. Sometimes we can redefine an industry without considering customers’ needs, but it doesn’t happen that often.

    Thanks again,
    Steve

  17. from James Saunby   /   November 28th, 2012 at 19:39

    Steve,
    Really interesting and accessible article. I come to this from the connectivity side – trying to estimate the impact IoT will have on connectivity requirements. It seems there is going to be a quite sudden move as more and more people understand the potential and develop the applications. At the moment it seems it’s driven by people with the skills trying to think what problems to solve – when the people with the problems start to understand how IoT can solve them, there will be an escalation in development.
    I think the enabler for this will be trying to bring the two groups together – like Hackathons are trying to do now, though probably needs some more formality?
    James

  18. from Steve Hilton   /   November 30th, 2012 at 22:43

    Hi James,

    Thanks for the comment. Some sort of Hackathon would be a great idea.Put suppliers and buyers together and hash out some of the challenges. I like it.

    When you say, “connectivity”, what do you mean? WAN connectivity (2G, 3G, 4G mobile, fixed-line)? LAN connectivity (Ethernet, short-range in office WiFi, etc.)? What do you think needs to happen at the connectivity layer to make IoT a reality?

    Thanks again,
    Steve

  19. from Iris Heilmann   /   August 21st, 2013 at 08:35

    Steve, it’s been a while since you wrote this blog piece. Still worth reading in our fast moving world ;-) and a very, very good and useful introduction. I personally think providers are already in between the M2M and IoT world – although reality and marketing sometimes differ. However, since you wrote that piece we saw a lot of discussion about security and surveillance, especially here in Germany. This discussion also nurtured the debate on the downsides and risks of such things as M2M, IoT, Cloud and Big Data. And I still believe that trust and the feeling that you don’t give up control is key for the broad acceptance of these technologies (although they are here to stay anyway). And we need this acceptance to make good use of the exciting opportunities which the IoT has to offer. So what do you tell people – end users as well as companies- when they are hesitant to enter the IoT, Cloud, Big Data word because of such thoughts? Thanks and looking forward to your answer, Iris