IoT (Internet Of Things)

  

 

 

 

When to come ?

 

Even though I don't agree with exact time scale of the curve, but I am a strong believer of Hype Curve (Gartner) in terms of overall life cycle of emergying technologies. (If you are new to Hyper Curve, you can easily get explanation and tutorials on the web).

According to my personal experience/observation on USB, Bluetooth, SmartPhone .. all of these has followed this curve.

It is also a good stuff to kill time.. pick any item or industry that you personally are interested and google the path on the curve.

 

In summary,

  • IoT was completely gone in Gartner's emerging technology hype cycle since 2019
  • Google Trend shows the level off and gradual decrease since 2018.
  • Number of IoT related YouTube I can easily found has decreased drastically since 2019.    

Does it mean that IoT is already dead ? or so mature that it is not classified as an emerging technology any more ? I don't think it is dead nor mature. It looks abvious that the IoT story does not seem to go through as the initial visionaries hoped/expected, but I think they are slowly growing in the background starting from various home automation / surveilance systems and various other household electronics over WiFi. I am not sure when IoT over Cellular or IoT specific Wireless technology will spread as early visinaries hoped, but I think we will see gradual growth of IoT over WiFi will continue for next few years to come.

 

 

 

Hype Cycle

 

Following is the Hype Curve for IoT that I found and I will keep updating as I find the latest.

 

2011
 

2011 is the first year when IoT shows up in the curve.

It was estimated to take more than 10 years before it reaches Plateau of Productivity
2012
 

Almost no change in terms of IoT.

It was estimated to take more than 10 years before it reaches Plateau of Productivity
2013
 

IoT picks up momentum and approaching towards the peak.

It was estimated to take more than 10 years before it reaches Plateau of Productivity
2014
  IoT reaches the peak (not a good news), but it is estimated to take '5 to 10 years' to reach the Plateu of Productivity
2015
 

According to 2014 Hypercurve, IoT would have been expected to be a little sliding down, but it still stays on peak of the curve.

Would it stay continusly at the peak and directly jump to Plateau ? or start sliding down gradually ?

Note : A new word 'IoT Platform' is registered on the curve.
2016
 

Interestingly, the keyword 'Internet of Things' completely disappeared.

What does it mean ?

 

The term 'IoT Platform' that registered last year is approaching the peak.

 

 

 
2017
  Similar trend to 2016 except that IoT Platform got closer to the peak.
2018
  Now IoT Platform is over the peak.
2019
 

IoT Related Technology is completely out of the curve.

Does it mean that IoT is now matured technology ? or at the verge of extinction ?
     

 

 

 

Google Trend

 

Another way to foresee some near future would be to check Google Trend. Following is what I checked in Google Trend. The tricky thing is to figure out proper keywords that can best capture what you are interested in. IoT case this is even more difficult since there are too many words hovering around directly or indirectly indicating this area (e.g, IoT, Internet Of Things, Machine to Machine Communication, M2M, MTC etc). I just picked the two keywords that I think are the most common words and got the trend chart as below.

 

 

 

 

 

 

 

 

 

Is it good timing, Now, to jump in ?

 

As of Now (Jun 2015), IoT is not at a good spot on the curve since it was already at the peak a year ago and I might have starting sliding down into long/dark tunnel of 'Trough of Disillusion'.

==> Adding Comments on Jan 2020. It seems the situation hasn't went very well as I described below (about 4 year and half year ago). We still don't see any matured IoT industry even now. See Why So Slower than Expected.

NOTE : Please take my comments below as a reference. As of now (Nov 2021), It doesn't seem that my expectation / assumption was right around 6 years ago, but I decided to leave my original comment as it is just to track down my own line of thinking over time.

 

If you are front end marketing/sales people who needs to make money right away, it would be not a good timing to step in because you would see almost nobody talking about this even before you have any tangible things to sell. But if you are on more strategic/long term marketing, it would not be a bad timing to get in. In many technology, most of R&T starts during the time when there is no hype/no fancy words/no buzzwords. It would be good timing to search and establish the relationship for the potential customer. By the nature of IoT architecture/eco-system, it would be highly likely that it is going to be 'long tail' market rather than conventional '80-20/90-10' market. So finding smaller but more customer will be extremely important in terms of product marketing.

 

If you are a decision maker on the investment in a company, definitely you may want to wait until the mid/end of Slope of Enlightenment. I would try the same if I am a decision maker because you can minimize the risk for the investment. But it should not be like that for several reasons.

    i) If every company decide to wait until that point, the industry and ecosystem will never form and no chance to sell forever. Of course, the best strategy would be to let other take all the risk and you will get in the market when everything is clear... but that would be the area of high level game theory ..  finding Nash Equilibrium ? :).

     

    ii) There is another reason why the decision maker should not wait too long. As you might have experienced, no technology can be implemented overnight however simple it sound like. If you start too late, you will be struggling for selling unmature product in the market and end up losing the whole market by the time your implementation is mature.

 

If you are a R&D person and your company decided to chase this market, I think this is the best timing. This (following a couple of years) would be only period during which you calmly sit in front of desk and doing a real R&D without being disdurbed by Sales/Marking section saying 'We need this, We need that.. my big customer want this and the deadline is yesterday' . No offense for Sales/Marketing people. I know you are the person who bring in the bread and buffer for my survival :)

 

My personal opinion ?  I agree with the hyper curve, but I am a little bit more optimistic in terms of time scale. Considering the accelerating nature of technical development, we may see a little early see the next phase than the current curve estimate.

 

 

In the past several month (as of Mar 2017), I have seen several promosing events and trends that make me think 'I am not sure exactly when we see IoT as many dreamers are talking about, but at least it seems that something tangible (touchable) trends are happening in this area'. Those things that attracted my attention were

  • Softbank aquired ARM at one of the largest investment
  • Qualcomm aquired NXP
  • The demand in Memory chip reached another peak
  • So many different types of embedded board and modules are available at very low price (You may easily get an embedded board to play with.. at a price of a fastfood meal).

However there are still many things that has a pretty long way to go (just in my personal opinion). Followings are some of these list

  • IoT module based on Cellular technology is still too expensive (When I say 'expensive', it means the price for single quantity purchase is too high). In case of GSM module, you may get it under 20 USD. However, you would need to spend at least over 50 ~100 USD for 3G. LTE based IoT module is not even readily available.
  • Even when I have IoT end-device, it is still hard to find IoT platform that are readily available with super low cost or free and also with very easy interface.

 

 

 

Why so slower than expected ?

 

As of now (Jan 2020), we don't see any hypes or talks about IoT as before but it is obvious that IoT is not yet in the matured state of the industry. Just as a simple indication, look through the series of Gartner's Hypecurve. IoT started appearing in 2011 and stayed there until 2017.. then appeared as a little bit different name in 2018 .. finally completely disappeared in 2019.

Why ? Let's think again about a few key claims that had often been mentioned about IoT devices.

  • The device will become much simpler ==> May be true for hardware, but in terms of radio protocol it became more complicated mainly due to reptitive transmission. Scheduling has become more complicated and testing/verification was much more tricky.
  • The IoT chipset should be very cheap ==> It is cheaper than high end cellular modem chip, but far away from the expected price target. This type of low cost chipset does not give strong motivation for main stream modem maker. At early stage, it attracted some of small scale modem development company but the cost/verification of the development was not as low as expected. Simply put, it is still too much in terms of development for small scale modem development company and too little in terms of business for main stream modem maker.
  • Special USIM (eSIM) that is expected to be installed once and can be profiled wirelessly regardless of where the device is located. ==> Is this turned out to be true ? Technically possible and really adopted by some device, but mostly adopted by highend cellular phone rather than the low end IoT device.
  • Battery in the IoT device should last over 10 years without recharging. ==> I think this is far, far, far away from being true as of now.
  • How about non-celluar LPWN chips like LoRa, SigFox, ZigBee etc ? ==> I think it will be very difficult to get wide coverage unless there is super strong demand in the industry.
  • Other reasons mentioned in here.

 

 

 

Cellular vs WiFi

 

Which radio technology would be the major technology for IoT ?  WiFi ? or Cellular ? or IoT specific technologies like LoRa, SigFox, ZigBee ? At least for now (Oct 2021), the answer seems to be obvious (at least to me). I think it will be WiFi for many years from now. The advantage of WiFi over other technology can be listed as follows.

  • The price of WiFi Chipset/Module is much cheapter than the chipsets for other technology. I think it would take only a few dollars for WiFi module and the module for cellular would be at least 10 times more expensive.  
  • Engineering cost of developing IoT device with WiFi would be much cheaper than other technology. You may test the IoT device with low cost spectrum analyzer, regular Access Point, WiFi Snipper and Wireshark to test your device whereas you would need super expensive equipment for the cellular based device. On top of it, it would be easier and less expensive to find/hire engineers for WiFi than for cellular.
  • With WiFi IoT device you don't have to worry about Gateway since any regular WiFi access point can work as the Gateway for the IoT system whereas you need expensive base station around your IoT device.
  • No additional cost for the service whereas you would pay for subscription fee for cellular device.

Of course, nothing is perfect and there are drawbacks with WiFi as listed below.

  • Narrower coverage. Usually WiFi Access Point has narrower signal radius (e.g, just a few tens of meters in radius) whereas cellular basestation can cover from a few hundreds to several tens of kilometers.
  • No or Poor Mobility. WiFi is very poor at handling mobility situation comparing to cellular system.
  • Poor Security. I am not trying to say WiFi Security is poor by itself. I mean it is generally considered as less secure than the security system in cellular.

 

I think at least for now (Oct 2021), the advantage of WiFi based IoT devices outweights the disadvantage.  Cellular people has been making efforts to put emphasis on the list of WiFi drawbacks because most of the WiFi drawbacks are resolved in celluar by design. However the question is 'How much of IoT application are really concerned about those drawbacks ? How much of IoT manufacturer or end-user are willing to pay extra cost to switch to Cellular in order to remove the WiFi drawback ?

 

 

 

Reference

 

 

YouTube