REGISTER | LOGIN
Breaking News
Blog

IoT's Assumptions Trouble Me

View Comments: Newest First | Oldest First | Threaded View
Page 1 / 3   >   >>
spike_johan
User Rank
Author
Re: All good comments but...
spike_johan   5/11/2016 5:30:10 PM
NO RATINGS
@sranje

It's not a matter of who exactly but more a matter of what I've seen in some of the IoT architectures that certain vendors have included in their prereleases.

Some of those architectures have been so spec in nature - as in generic, as in build a house on spec - that the implication is their presumed architectures are universal and genderless (aka one size fits all).

Example. The presumption in some architectures is that all sensors everywhere should be intelligent. That is just plain wrong for a couple of different reasons: One, an intelligent sensor would require some version of a protocol stack and  processing power, making it more vulnerable to hacks, not to mention making it way more expensive than its less intelligent brother.

Second, for many sensors in many similar applications like measuring simple values like temperature, pressure, and volume; added intelligence does not add any such benefit when a secure, upstream intelligent controller can be designed to provide the control functions for dozens if not hundreds of dumb sensors.

Dumb sensors managed by intelligent controllers was once the design paradigm used in the petro-chem industry back when I was just starting out in the late '70s, early '80s. And that solution worked just find. (And you couldn't remotely hack into something like a thermocoupling.) And I reckon that the dumb sensor/intelligent controller model still has many applications in some of today's modern industries.

But back to your question and to use another example - unless I am misreading Intel's IoT literature (who is in the processor selling business), their IoT strategy seems to be very processor centric. And it appears to me that if they have their way all sensors will either be coupled or embedded with some version of their Quark series processor.

But, hey that's just my opinion.

 

sranje
User Rank
Author
Re: All good comments but...
sranje   5/11/2016 1:24:17 PM
NO RATINGS
Hi Spike --- who exactly is still making such homogeneous assumptions?   ;-))

spike_johan
User Rank
Author
All good comments but...
spike_johan   5/9/2016 5:35:35 PM
NO RATINGS
The biggest gripe I have with everything I've read so far about the emerging IoT is the assumption that the IoT is a one size fits all proposition.

The reality is that every sector in every industry - commericial, consumer, industrial - all have different requirements concerning the monitoring, reporting and management of its data.

As an old EE, I've worked in petro-chem, telecomm, and software development during the course of my career. And from what I've seen the only thing those three industries have in common from a data perspective is how they store it.

Everything thing else - from both the hardware and software side - including the processing and transmission of that data is different.

sranje
User Rank
Author
Re: One line -- Excellent comment !!
sranje   5/9/2016 2:14:18 PM
NO RATINGS
Thank you Sheepdoll -- for a wonderful extensive commentary  !!

Although I disagree with you and Rick -- I find, for example, my Watch useful and increasingly so.  

And -- most already outstanding IoT applications are -- not Consumer at all

Patrick Mannion
User Rank
Author
Re: It's just a (temporary) term
Patrick Mannion   5/9/2016 11:55:16 AM
NO RATINGS
At a higher level, what IoT brings is the layer of analysis of data of the connected nodes, and it's that which we're just beginning to understand, appreciate and apply. The assumptions are big, but so too are the challenges and rewards as it unfolds.

 

Patrick Mannion
User Rank
Author
It's just a (temporary) term
Patrick Mannion   5/9/2016 10:27:07 AM
NO RATINGS
The IoT is just a new name for what's been happening -- and will continue to happen: connecting everything. As for the FitBit, well. Who knows. For many it's just a reminder of what they're not doing so it can be demoralizing. My "smart" phone was telling me on Saturday that I needed to take some more steps, so I gave it to my daugther to hold during her step-dance practice. I went from 1000 to 3000 steps in about 20 minutes. I could feel the pounds melting away with every click and clack of the shoes on the floor. Felt so much fitter -- the smartphone told me so!

alex_m1
User Rank
Author
Re: Fully agree
alex_m1   5/8/2016 9:08:58 PM
NO RATINGS
The cat idea is a good idea(unless you're a dog). But not an easy one, mostly because you need to convert everybody at the same time, you'll need to ensure a high level of cat safety, but zero increase in risk for humans(because of attentional issues of drivers, and the need to convince car dashboard companies). Maybe the winning solution would be just to put a radar on chip, on the cat and if before crossing a road there's a car at the distance, use something annoying sound to guide him off, like invisiblefence.com ?


And sure, there's no need for a killer app. But i think there's already a many killer apps - in usinesses . They love to optimize stuff, and the IOT can be very helpful at that. And consumer product could benefit from the tech.

Jayna Sheats
User Rank
Author
Re: Fully agree
Jayna Sheats   5/8/2016 8:22:38 PM
NO RATINGS
We are clearly in sharp agreement...

Here is an example of what will, I am confident, emerge from the very dusty current conditions.  It is very important that I not let my two cats outside lest they be hit by a car (the danger in the neighborhood is very low, but it happened to a previous cat and for very personal reasons that can't really be explained here, this is a chance that neither myself or my son is willing to take).  Of course cats love to be outside and explore.  So what if I could buy a tag at the pet store which, when I put it on the cat's collar (it can't be bigger than the usual collar tag) and activate it, will send a signal such that if it is in the street, any car that drives down the block will get a warning on the display (which is already a part of all new cars) in time to avoid any collision? 


I am not the only pet owner in the country by a long shot who would pay decent money for such a thing.  Of course it has to "just work" in a true plug and play fashion, and we are very far indeed from being able to do that (otherwise I could probably get a Silicon Valley VC to give me several M$ to try it...).  But that is really the point: waiting for the "killer app" is a recipe for paralysis (as in the aphorism of physicists about math (too much rigor leads to rigor mortis).  The frivolous and the not-very-useful (as in the fitness tracker that started this conversation) will "pave the way" (a horribly over-used phrase in the press these days - every single technical innovation, however minor, seems to be in the paving business) for more useful versions.

alex_m1
User Rank
Author
Re: Fully agree
alex_m1   5/7/2016 6:07:17 PM
NO RATINGS
Jayna, i agree with what you said:experimentation but also simplicity, hidden, etc - i.e. very cheap in effort.

But the weigh example point to another issue: adding IOT doesn't add a lot of value, and often times the value of most data is very low(telling people to lose weight doesn't work, even from a doctor - heck medicine probably doesn't even have a diet regime that works long term, and unless you have some specific diseases, you don't need accurate and fast weight monitoring) , and often the data is only valuable in aggregate over many sensors. Often those leads to a chicken/egg issues - it's only valuable to build the tools for doctor when everybody has connected scales, but than why have connected scales ?


So it leads to the conclusion that the IOT also need to be very cheap in monetary costs to reach it's potential. But when you look at the pricing of connected devices - it seems out of touch with value(at least for non-wealthy consumers), especially if you compare it to other electronics like smartphones. 

The company to watch here would be xiaomi , i think, since they have a plan to offer many connected products at very affordable prices(while having great marketing), and so far they seem to do it. For example, their smart wristband costs around $15, or their smart scale cost $18 which is about $5 over a cheap analog scale.

Sheepdoll
User Rank
Author
Re: One line
Sheepdoll   5/7/2016 3:16:52 PM
NO RATINGS
@pixie1 you nailed it.

I sort of stopped attending trade shows, when I attended a lecture who was showing IoT connected bras (<- Max favorite symbol here.)

In reaction to this I put in proposals for another IoT based trade show. This was rejected out of hand.   I think I came across too much as cassandra.

The talk was to be on the 500 year to 1000 year history of watch design.  Titled Failure through success.  Companies like Breguet, Waltham, Elgin and other "famous" brands.  These companies had 100s of years of success before their retirement rolls outdid their profits.  Some of these brands would dormat for 100s of years.  What I was going to show through design was how designs of the 1790s looked new in the 1930s.  Design aesthetics cycle.

Designs also cross genders.  The ubiquitous wrist watch.  Queen Elizabeth I sometime in the 16th century was given a bracelet with a watch embedded into it.  Wristwatches did not become in use until WWI.  Why?

A company name Omega made a Ladies dress watch that was more reliable that most of the others.  These were clipped on to one's blouce they also had the dial upside down to be easier to read.  Some of these watches had holes drilled through the cover so as to see the time.   Girlfriends would send said watch to their soldier boyfriend as a memento.   Clipping the watch onto the tunic was not practical.  It was found that by soldering crude lugs onto the watch case it could be tied to the wrist.

The other thing that happened in the 1880s was the discovery of radium.  This allowed the hands to glow in the dark. It was also discovered in the 1890s and early 20th century that the women who painted said dials, tended to get sick.  They would lick the brushes with their tongue to point it.  This happened again in the 1950s but the stuff then was Strontium-90. Another solution still looking for a problem.  Lessens not learned.   Most watches now use Tritium.  At the 6 oclock position next to the country of origin there will be a letter telling which radioactive isotope makes the hands glow. 

These two things combined to make a product intended for women to be adapted by men.  By WWII most men dropped the farmers turnip (as the pocket watch was called) in favor of the wrist watch.

These watches were designed to be serviced annually.  This way the jewelry industry had a guaranteed visit by the "dad"  I have some trad magazines from the 1940s 1950s that tell how when dad comes in for the annual watch cleaning to sell a necklace to "mom" a class ring to "junior" and a tennis bracelet to "sis."

This in the 1960s was disrupted by the surplus of mechanical bomb timers.  A Canadian company called United States Time had an overrun.  These  were put into watch cases and sold for a fraction of the cost and were disposable given that they were riveted together.  Jewelers would not sell them.  So the sales reps went to drug stores and created the "drug store watch."

A clever marketing campaign with things like elephants stamping on them and the slogan "It takes a liking and keeps on ticking" made at the "Timex" a household name.

Our industry ,Notably Ti, But all the calculator manufactures had one too, crushed Timex amd 1/3 the swiss economy in the 1970s with the "digital watch"  I still have a Fairchild watch.  Would love to find a Nat Semi watch.  Even then this was a solution looking for a problem.  I still remembering one of my engineering mentors complaining that one had to press a button to see the time.  That this could not be done while driving.  Pretty astute for someone in the 1970s

Omega (who really did invent the wristwatch in WWI) countered by re-branding itself as Swatch and made a hybrid electronic mechanical watch (and gave us SMT processes based on 8mm film processing equipment.)  SMT according to one of my trade magazines originally stood for Swiss Mounting Technology.  Although the acronym would more properly be something like CMH. 

One of my regrets is that I never got an apple watch from the 1980s.  These were promotional items that had the apple logo at the 12 o'clock.  They came in both the rainbow apple and the single translucent one in the 1990s.

Apple's own digital IoT watch evolved out of the iPod Nano that had skins to display the time (including Mickey) and aftermarket writs bands.


On the other hand I recently got an Annikken Andee and am connecting it to the front panel of the pipe organ.  This way I can easily replace the front panel with an iPhone or iPad over BLE. 

The most professional (and expensive pipe organ relays) have for the last 7 or 8 years required a router so it is not much of a step to connect them to a cloud or other such service.  Not because there is any need to.  It is simply something that could be done.

 

Page 1 / 3   >   >>

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed