Thread is a mesh network to connect products around the home
threadgroup.org
threadgroup.org
The strongest push for IoT seems to be from the wireless hardware manufacturers (Cisco, Qualcomm, etc.) and of course, they have the most to benefit from putting a wireless chip in everything. It's a wonderful solution to their problem of "How do we sell more chips?"
Now Google and Apple are hopping on board. They also have something to gain; Apple sells more hardware, Google collects more information.
But from a consumer's perspective, it seems like a solution in search of problem, and I haven't yet seen a single truly compelling use case that will impact a large segment of the population (turning heating/AC or lights on and off automatically is cool, but not that cool for most people when initial cost is factored in). Can somebody illuminate this for me?
By far the most useful is the Z-Wave deadbolt on my front door. It means I never carry keys, I can give codes to friends or the contractor and get texted when they enter, I can set temporary codes for deliveries, etc. I never have to worry if I locked the door.
I was surprised at the hype around Lockitron, since these have existed for a while. The main different seems to be that people prefer WiFi (since they already have it) even though it's a poor protocol for low-power and low-bandwidth.
ie, I think it is actually too early in the IoT life cycle to really visualise the impact it will have.
Let me explain - currently our lighting and applicances are designed entirely around the idea that they will be manually operated and standalone pieces of equipment.
Once that assumption is taken away, and houses are built with a different assumption, suddenly magic begins to happen - imagine that instead of a few lights in the lounge, there was a ceiling and walls covered in thousands of LED lights that could be independently controlled via the internet. The average home could literally have lighting shows. An entire wall could be utilised as alternately a television or a photo viewer, even a wall in every room.
etc, etc. Use your imagination here. The point is that the impact of an IoT on the average middle class house as they are built today is currently pretty small because neither the house nor the contents of the house were designed with an IoT in mind. Once we actually have an IoT houses will be build from the ground up with different assumptions .
At least, that is my opinion.
tl;dr: The IoT has a chicken and egg problem.
The question the parent was asking was "Why hasn't anyone shown me a concept that makes me think, 'Now I want an internet of things in MY house!'?" Not, "Why haven't I seen quick adoption of these IoT concepts out there?"
People do - "mood lighting" is already a thing, that is why we have dimmers on some light switches now. Going further than that we can change lighting color, create patterns etc.
Many people, including myself and my kids, would LOVE to have that in their houses.
But only if it was easy, it needs to come essentially for nothing, ie, to be builtin from the start.
"What does that even have to do with the internet?"
Internet Of Things I have always understood to mean "Things that we can talk to and control using our devices, possibly over the internet" but in terms of the internet as it is widely understood, not much really. I am not sure I understand the question?
"We already have huge televisions that aren't very expensive. If we want the whole wall to be a TV, we can get a projector. A wall covered in individually-addressable LEDs is probably not going to be cheaper than a wall covered in televisions or just a big projector."
Right, and who knows whether that specific leap of imagination is realistic, I certainly do not - but I was not trying to make you want a TV in your wall, or even arguing that it specifically was a good idea, it was merely one possible example to make a point.
"The question the parent was asking was "Why hasn't anyone shown me a concept that makes me think, 'Now I want an internet of things in MY house!'?"
Yes. and I was attempting to suggest that one possible reason is that our imagination is balking because we keep looking at what we have in our houses now - all of which is extremely functional and works fine according to our expectations and desires, which have been moulded by what is available and by what has always been available.
ie, we are looking at our existing house infrastructure, the idea of an IoT and saying "but the Internet Of Things is not going to give me a faster horse, so why would I want it?".
I think I am doing a bad job of explaining myself, so I will bow out :)
I think again you're missing the point with this. An LED light show is a trifle, it's not something that will really affect you in any way and it's already possible with current technology. A cell-phone controlled LED light show is the faster horse. What the parent (and now, by extension, I) was saying was that I'm looking at the way an inter-connected house would look and I'm not seeing anything fundamentally new, just marginal and possibly dubious improvements on our existing technologies.
With Nest it's obvious that the appeal is that you'd have the ability to remotely change your thermostat settings, which is neat, but how often does that really come up? If I were to program my thermostat with my work schedule generally, how often would I really need the internet-connectivity? Once a month maybe? Is that really worth it? It's an improvement, but it doesn't seem like it's a whole new platform like you and other people are suggesting. Not yet, anyway.
>Internet Of Things I have always understood to mean "Things that we can talk to and control using our devices, possibly over the internet" but in terms of the internet as it is widely understood, not much really. I am not sure I understand the question?
I was mostly referring to the LEDs-wall, thing, which isn't inherently a networking problem any more than addressing each pixel in a television is a networking problem. If you're building a wall of LEDs for "light shows", chances are you'll just connect them all to a single controller anyway which can individually address each LED "pixel", and then that controller will be either plugged into your "internet of things" or it will be controlled by some physical panel.
The two big advantages of device connectivity are going to be that you don't have to be in the same room as your device to mess with it (not having to get up basically doesn't count, because we already have remote controls for almost everything), and that your interface with a device can be much, much more robust (you can essentially use your tablet or phone to give a full touch-screen UX to a small, cheap device like a coffee maker or a lightbulb). The first advantage seems to be getting a lot of hype because people can mess with things at home from when they are out, but that's a benefit that I find marginal at best (how often do I really need to interact with physical things in my home when I'm not even there). The second benefit is much more valuable, but again it is likely to be marginal, because chances are your major devices are ones where you don't actually need to change the settings (or replace parts, etc) very often, so it's a convenience maybe 2 or 3 times a year, not a revolution in the way we interact with our homes and devices.
And likewise, do we need an Internet connection in order for the object to be smart? If context awareness is the name of the game, there's quite a bit you can glean by being a physical object with a memory of how it's used, and beyond that, a local network.
The Drift lightbulb is interesting because it adds some smarts to a dumb object but doesn't add more I/O - it still uses a light switch and an LED bulb, and it isn't trying to get on the network. There are computers that are taking on the form factor of a light bulb, and there are light bulbs that are taking on the force multiplier of a computing device. The latter philosophy needs to be explored more.
The magic of a tenfold increase in maintenance costs.
I think the company to look at is Philips. They really understand how to create appealing connected things. For example this article[1] that talks about their controllable color lights, cooking assistant and coffee maker, all controlled through an app to great benefit.
And once something is connected, bluetooth becomes a limitation - say you want to choose coffee type at the living room with your guests ? or you want to watch the cooking assistant while being at the kids room ? Considering that the cost of full connection should be similar to bluetooth in the best case scenario , why settle ?
[1]http://gigaom.com/2014/04/04/lights-cookers-and-coffee-a-dee...
No mesh networking? That has been in the pipeline for quite some time now, and will hit the market big time:
http://www.csr.com/news/pr/2014/csr-mesh
Disclaimer: I've been part of a Dutch consortium with companies that developed their own mesh standards and did not understand that you need to "board the smartphone train". This is why I'm skeptical w.r.t. anything that does not just interface with smartphone technology.
Another area seems to be from those offering yet another (non-standardized) protocol or format such that they can then sell you tools and consulting tailored to their software.
[I]t seems like a solution in search of problem ...
Indeed. Most of the demos I've seen (e.g. turn on a light using Twitter) are easily doable in a number of ways, so the requirement for some IoT "solution" is dubious.
But suppose we grant that such demos are mainly done because they are easy to comprehend, easy to show in a bite-size video clip. Still, there still needs to be something that extrapolates to a more significant use case. I'm not so sure these currently exist for most homes.
Part of me thinks people are tagging "IoT" onto anything they can because it sounds sexy. Controlling my living room lights from my phone is not an example of the Internet of Things.
Controlling fertilization for a farm based on smart sensors out in the fields? That I can see.
Actual use-cases in the average home are less obvious, as is why people will want it.
Keep in mind that in terms of factoring in costs, yeah, that will be true at first. But it wasn't that long ago that you had to buy an aftermarket modem as part of the barrier to entry to getting online on your personal computer. In a few years, connectivity will be the new normal, and prices will adjust.
No doubt the costs will come down, and capabilities will increase. And as costs and capabilities increase, potential for innovation will increase.
But it would feel much more compelling if they had a strong application out of the gate (perhaps its networked thermostats). It would benefit from some initial momentum. Industry seems to be taking a "if we build it, they will come" approach, but this often doesn't work out in reality.
Is that a tech prediction, or an optimistic view of your future mental health ?
Kind of chilling, actually: The notion that this technology might make us more complacent with our weaknesses rather than empower us to be better.
On the other, if you can remotely control these devices then I have to wonder about the security implications. Track records for most wireless device companies have not been good, and I don't want someone burning down my house.
Maybe the plan here is to get everyone on a single standard that shoves good security practices down their throats? But you still have to rely on them implementing it correctly.
From another perspective, front doors with deadbolts and residential don't have a good security track record, either, since most of them can be kicked open easily. But they're good enough.
Shameless plug: http://www.flair.zone We are aiming at lowering HVAC energy consumption by between 20 and 30%. Financial incentives are rather attractive ones.
I will agree that there are going to be IoT whizbangs that are solutions in search of a problem, but there are bound to be a handful of actual problems that get addressed.
I'm personally waiting for my refrigerator to detect my food inventory, and auto-order grocery delivery once a week to restock itself. But seeing how Safeway, Shopping Express, Amazon Fresh, etc. don't even have an API yet, it's going to be a long wait.
I imagined this as well, and it would be pretty cool (although I actually enjoy grocery shopping).
As you say, it's still a long way off. In addition to Internet connectivity and an API, fruit/vegetables/meat and all other refrigerated items would need some kind of tracking label (RFID perhaps) so that the refrigerator could automatically track levels. Scheduling delivery of refrigerated food might also be an issue (I'm not sure how existing online grocery delivery services handle this), as I don't want my dairy items sitting on my hot doorstep all day waiting for me to get home from work.
This could be extended to pantry food items as well.
Problem is the biodegradable RFID tags don't exist. A dutch company was researching them a few years ago but that ship sailed when the iPhone 5 came out without NFC.
Google's project tango could be used with a standard library of objects, but that's still too early to tell if it's possible. Not to mention you'd be tied to Google's services and their whims.
Instead of using RFID, it's using barcodes. It's a little more manual, but same basic concept.
However, its coming, much like the cloud has.
Three examples:
I was recently looking for a door chime ("door bell") and since I have WiFi already all over the house I wanted something that could piggy-back off of that to increase range but also maybe send the notification to my phone or similar. Doesn't exist. Only ye-olde style RF doorbells no different than what they sold 50 years ago.
I wanted a motion sensor that wasn't part of a security system, just a single stand alone motion detector which could send a text message ONLY during set hours (e.g. 11 pm to 6 am) when motion was detected. But nope, also doesn't exist (in fact there is a serious lack of variety in motion detectors, the closest I found was a Honeywell one which set off a doorbell when motion was detected, but cannot be shut off based on time of day).
I want to know when the clothing dryer is done. I often don't hear it. I have WiFi, but yet I cannot receive an update digitally yet. Why?
I cannot speak to everyone but for me personally I WANT this stuff. I'd happily buy this stuff. It doesn't exist. Where is it? Make it and I'll buy it.
Here is how I see it coming together:
- STANDARDS: The "message bus" needs to have essentially an HTTP protocol standard, that both producers and receivers on both ends can understand.
- Hardware, appliances and objects generate their own implementation of an end-node adhering to the protocol.
- Wifi Routers allow the firmware to run "master" nodes, that would have a mobile app and a web front end to configure and customize.
But because having to VPN into your house to manage anything, or even have to have a VPN at all would be silly. So now we host it in the cloud and the devices simply pull.
To design the APIs, platform, SDK, and hosting solution would look a lot like what Apple and Google are doing with their all-encompassing themed keynotes this year. I'm actually worried about interoperability more than anything.
I think a better analogy to what we need should be an HTTPS protocol standard. HTTP was designed in an era of slow computers when security and privacy were not particularly high on the list of developer priorities, and now we're seeing an agonizingly slow move towards universal HTTPS adoption. It'd be nice to actually start out with security-aware protocols for once.
For every "cloud," there are many overhyped technologies that went nowhere, despite industry's best efforts to thrust them upon us.
Bureaucracies and committees are good at developing protocol specifications. Instead of throwing spaghetti at the wall maybe those of us interested in pervasive network computing should get an ANSI committee (or IETF-like task force) together to develop the de-facto standard language.
I don't want to have to build out an API in front of my cluster that speaks every possible permutation of language from every device and sensor. As a hardware vendor I wouldn't want to lock myself out of an ecosystem because my device only speaks a particular language that only a portion of the cloud speaks (or may become obsolete in a few years).
It's nice that this stuff is happening... but a Cambrian explosion of protocols isn't going to make the IoT very robust, IMHO.
Cool that it maps to HTTP / REST so you can proxy it to the web. All the tricks it gets from UDP + TURN ICE STUN are p2p friendly.
However, also as with OSC, you can use TCP for CoAP if need be.
However, I'm concerned about the openness of it. I'm like 99% certain that the only realistic way to get a system like this to work is a 100% open, unlicensed, freely-implementable protocol with absolutely no restrictions. It doesn't currently sound like that will be the case, which is going to be a problem.
Perhaps I'll be working on this after all…
From what I read, it doesn't give you things like send command X to dim lights. Zigbee HA already defines those but everyone violates them in some way.
Here's hoping it dies off quickly.
RAND-RF = royalty free
If I were to bet, I'd say it's a way of guaranteeing they retain external access to in-home data. Just market it as ensuring interoperability and the apologists will be coming out of the woodwork.
I assume it's so you don't have thousands of non-compliant devices being sold with the same name on them. Avoids the "Oh no, that's spec+- not spec-+, they're not compatible unless you've got a bridge with a chipset with part ID starting with 71204, 71205xxx are not compatible".
As I've said elsewhere on this thread, I hope this dies, and it's yet another nail in the coffin for Google's once-deserved reputation for supporting open ecosystems.
I'm sure the reason they have membership, unlike the rest of this thread, is that they need someone capable of signing a patent cross-license to keep it royalty free.
I agree that there may be good reasons for this, but the existence of those reasons doesn't make the protocol any more open.
What?
There are plenty of open protocols produced by membership based groups.
I believe you have redefined open to also include some notion of governance, which, while popular, was never the case previously :)
If you want to define what you mean by open protocol, that would be helpful, since everyone seems to have a different definition nowadays ...
Page 2 of the presentation says it's supposed to be an open protocol.
"What will be the licensing and royalty terms of membership?
The Thread technology will be licensed to members under RAND-RF terms."
Yale is doing this because they have locks that already have Zigbee.
ARM is doing this to sell chips. They have their fingers in a million different protocols. I don't blame them.
Funny enough, there are very few things in home automation that actually need battery. Everything from light switches to light bulbs to smoke alarms and security systems are wired.
The better way would be to use BLE which has the same benefits of 802.15.4 stack. BLE 4.1 allows you to do mesh routing, like Zigbee. The biggest boon is that you can talk to the nodes directly with your phone.
When your lightbulbs become smart devices, they no longer want to be on an on/off circuit. In fact, it doesn't even matter if a group of them is on the same circuit or not. You just need a switch to talk to them wirelessly over Thread, ZigBee, Enocean, or similar. At that point, you don't have to run the circuits to your light switches either.
Enocean's switches are all powered by kinetic energy, but if you don't mind replacing all of your batteries every few years, that can work too.
Indeed, and sometimes can't be. There are a couple of locations in my home where there's insufficient depth to mount a wired switch and I instead have MK astral (z-wave) low-profile battery powered switches.
I think the better option is, as you said, have all switches be wireless, and somewhere like a basement and have a bunch of 'controllers' to tell it to switch on or off.
The one thing I wonder about is replacing lightbulbs when they fail. It's a good safety practice to kill the circuit when you're doing it, and without a hardwired switch you'll have to do that at the breaker now. On the other hand, if you're buying fancy wireless LED bulbs, they'd damn well better last longer than an incandescent A19.
BLE and 802.15.4 have near identical modulation, and BLE has very low overhead compared to Zigbee.
From a safety perspective, how the bulb is controlled is irrelevant.
I'm not too hopeful yet. Maybe it's just their ".aspx" website that has a proprietary flavor to it, when we terribly need an open standard here.
We do need an open standard, so each inventor just needs to implement their idea, & not spend 90% of their time building the whole supporting architecture, and trying to train the customers in the differences in the setup. The wearables/IoT ecosystem will continue to flounder until the equivalent of W3C arises.
The horse may have left the barn already. We have the IETF, and they have offered CoAP. (http://www.utwente.nl/ewi/dacs/Colloquium/archive/2010/collo...)
But the gold rush to claim a chunk of the IoT and wearables product space means a good many people and groups want to invent Yet Another Standard.
With so many out there (some good, some wonky) it wil be hard for any single organization to stand up and declare anything remotely like a de jure standard.
(For myself I'm working on software to at least bridge some of the more sensible ones.)
And yes, that gives off the same vibe. It's a vibe that says, "we're not really into open source". I say that in spite of Microsoft's recent -- and welcome -- forays into open source. Their technology is still used primarily among people and organizations who are not very welcoming to open source technology.
Even though I'm a web developer who _loves_ open source on the server side and has pushed for as many open source technologies as possible in my daily work as a web developer, I much prefer the experience of MS's office suite to many of the open source alternatives.
And that doesn't undermine any of the work I do as a programmer. My job is to pick the superior option and weigh pros and cons for a given set of circumstances, whether it's open source or closed source.
I'm happy being a user and advocate of open source technologies without having to be an absolutist who won't use anything else.
Good to see truth in advertising.
If TWC has their way I imagine this will go about as well as Google Wallet did, which was killed because all the mobile carriers wanted to promote their own version (formerly Isis), and had the power to do so (since they could prevent Google Wallet from working on their devices.
I think they don't care too much about the technology, they want to market the monitoring and connection services, which I guess means they will still choose whatever favors collecting those monthly fees.
But i'm probably missing something here, since i haven't heard of anybody doing this. I wonder what ?
If I want to turn the lights off and I'm at home, I just flick a switch.
"Extensive support for sleepy nodes operating for years using even on a single AA battery" - http://www.threadgroup.org/Technology.aspx
Why is there no easy way to report errors on websites these days?