Interview: Contiki OS Creator On Building the Internet of Things 45
angry tapir writes "Last year Adam Dunkels, the creator of the open source Contiki operating system, launched a startup to build tools for the 'Internet of Things'. This week his company, Thingsquare, is releasing an evaluation kit that lets people test drive their IoT system. I caught up with him to talk about Thingsquare's plans and the hype around the IoT."
please stop the term (Score:1, Insightful)
It's so hipster-ish or New Age-y. You know, and Internet... of Things! /pssst... this internet is already made up of things. Go find some coffee that cost less than $5 and come back to me.
What IoT is supposed to mean (Score:5, Informative)
Re:What IoT is supposed to mean (Score:5, Funny)
So what's a better buzzword for that?
Thingternet!
Re: (Score:3)
The impression I get from reading the featured article is that "Internet of things" refers to giving each electrical appliance a microcontroller to connect to the Internet so that the appliance's owner can manage it remotely. This has applications in street lighting and traffic signal automation, industrial automation, and smart distribution and metering of electric power. So what's a better buzzword for that?
'The Internet of Things' is basically a polite way of saying "Hey, SCADA is so easy that it should be a consumer product!".
Re: (Score:2)
'The Internet of Things' is basically a polite way of saying "Hey, SCADA is so easy that it should be a consumer product!".
Unfortunately, as per the article, they're not even talking about any existing automation standards. It's all about "we're going to have to create new standards". I like clean-sheet designs as much as the next guy, but they want us to take on a new OS and then design new standards for it and hey, what do we need them for?
Re: (Score:2)
Since SCADA, X10, etc. was originally designed with serial connections in mind and no security at all it's high time for something new. I like their idea, although I don't know enough about the people involved to know whether they're competent or trustable to be given the responsibility.
Re: (Score:2)
I'm not against some new standards, but they're acting like nobody has ever done any of this stuff before, which is a lot of balderdash.
Is it really an internet of things before the things are talking to one another directly?
Re: (Score:2)
Re: (Score:2)
Well lets expand on the idea then. I mean we've all seen parts of the internet of things but where is the unified approach they are talking about? Everyone doing their own thing is not creating an internet of things. Their point is the internet only works because every task talks the same fundamental protocol. I don't think anybody has EVER done something unified like that before.
Some very basic attempts have been made but they seem to have come and gone. Before Google Power there was no unified way to acce
Re: (Score:2)
Why do we need a buzzword?
Re: (Score:2)
Why do we need a buzzword?
Because Slashdot limits the length of story titles.
still just a dream (Score:2)
Re: (Score:2)
IMO a stupid dream. I cannot think of a single thing I have at home that I would want to control from work. I mean, if I forget to shut off the coffeepot or the living room lamp I'm not going to know I forgot until I get home. This "internet of things" seems pretty stupid to me.
Re: (Score:1)
Re: (Score:2)
The impression I get from reading the featured article is that "Internet of things" refers to giving each electrical appliance a microcontroller to connect to the Internet so that the appliance's owner can manage it remotely.
That's probably as good a guess at the interviewee's...
""We don't have a good notion of what the Internet of Things is yet,"
Yeah, thanks for that. So what you're basically saying is that it's a techno-fetishists' obsession with networking everything for the sake of it masquerading as a solution looking for a problem. Which ironically *will* cause problems [slashdot.org].
The benefits? Well, er... I guess it lets the Boys' Toys mob show off their IP-enabled, iPod app-controllable nasal hair trimmer to their mates until they get bored of it five minutes later and lose it down the back of th
Re: (Score:3)
The benefits? Well, er... I guess it lets the Boys' Toys mob show off their IP-enabled, iPod app-controllable nasal hair trimmer to their mates until they get bored of it five minutes later and lose it down the back of the sofa.
Electrical systems suck. They suck because they are dumb. Your car might tell you that you've got a signal light out, but it won't tell you which one (front or back.) This is because it's cheap. But at some point each light will have its own microcontroller and its own relay (solid state, whatever, bear with me) and it will know how it is performing and be able to estimate its future failure, like a hard disk. Well, hopefully better than that. And in your house, wouldn't it be nice if your light switches co
I had a cunning plan... (Score:2)
..You just reminded me of it.
In your garage, or wherever you park your car, you could have mirrors opposite the "four corners" of your car so that you could easily check your lights directly from the driving seat.
They wouldn't have to be big, expensive mirrors. And if you can't see, it means it's dark and none of your lights are working at all.
Re:What IoT is supposed to mean (Score:4, Interesting)
Internet of things is yet another buzzword applied to an older term: home automation. Similar to how cloud computing is a throwback to mainframe computing. Street lighting, traffic signal automation, industrial automation, smart distribution and metering of electric power falls under automation and SCADA. I don't see a point to remotely control a washer or toaster over the internet. Though a local system might be helpful with remote reporting.
Though there are some appealing ideas that can come of connected home appliances, the most obvious being text alerts of status. I would love it if my washer and dryer could text me when they are finished. Nothing sucks more then forgetting to dry your work clothes and then finding the wet, wrinkled, mildew smelling mess in the washer the next morning. Same with the dryer, wrinkled dry clothes are just as useless as wrinkled wet clothes. Those are the only two good examples I can come up with. Connected refrigerators aren't as appealing. I know how many eggs I have left and how much milk I have left as I open my refrigerator daily. Microwaves and toasters aren't that necessary either as they spend little time doing the cooking. Maybe oven or stoves could be timed and send an alert that they are done. Thermostats and AC units could be monitored for usage and remotely turned on/off. That is another good use.
Home monitoring of power usage might be useful for people looking to cut their usage. If you had solar it would be good if you could compare your home load to your banked kw/h and see the usage in real time. "Your solar system generated 50 kW/h today. Your current load is 10kW, you have 5 hours of free power left." Then you could look at your current appliance load and see what uses the most power. A breakdown chart of appliances and their monthly power consumption could help people determine if they could use that appliance less. Clothes dryers suck up a lot of power, maybe during the warmer months they could use a clothesline. Hell I dry my clothes indoors during the winter too, the dry air helps.
Instead of wireless, which makes integration easier, powerline networking would be just as appropriate. Most of those appliances are either hard wired in or plugged in. A router would need a wall plug adapter and ethernet cable or better yet, the routers power cable could pass through to an internal power line networking adapter. That or use Zigbee on the lower frequency ISM bands.
Things get cheap, then possible (Score:2)
[...] I don't see a point to remotely control a washer or toaster over the internet.
There's a sort of blindness that people have when they see what exists and can't imagine it being different. The creator of Babylon 5 once described seeing an old SF movie (Flash Gordon maybe?) where the crew had to abandon a space ship. They grabbed their laser blasters (handheld), anti-gravity belts (little box ona belt), and the portable radio, a giant box that needed two people to carry it. Because nobody knew what laser guns or anti-gravity belts look like, they could imagine science making them arbitr
Re: (Score:1)
Re: (Score:2)
Re:please stop the term (Score:4)
Re: (Score:2)
Re: (Score:2)
Whoa, slow down there Satan!
Re: (Score:2)
Why do people equate home automation with online?
I have often considered hacking my coffee machine to give it a microcontroller and an IP address with a little web interface. It could sync with a program I have on my computer that automates other things like the garden sprinklers so that the machine is warm when I wake up (a time which changes daily but none the less the computer knows because it also sets off the alarm clock). It takes 20min to warm up properly so as of late I've given up on a morning coff
content free summary (Score:2)
I wonder if it's worth clicking the link, since the summary was content free, and doesn't really tell you a think about what they are talking about, or why it would be interesting to anyone.
Nah. Probably another slashvertisement.
Re:content free summary (Score:4, Informative)
It's not worth clicking the link, because the article is content-free. It gives no specifics beyond names of people and companies. It has no technical information whatsoever.
What Are His Thoughts (Score:5, Interesting)
What are his thoughts on the absolute saturation of the 2.4GHz spectrum and the increasing load on the 5GHz spectrum.
The internet of things sounds nice, but the amount of devices pointlessly WiFi enabled and broadcasting in my home is having a very negative effect on the usability of the spectrum for anything of value. Does he have any plan to mitigate this growing issue? Does he care at all about it or is he solely focused on funding and an early exit with a fleeting; 'I'm rich, bitch'?
Re: (Score:3)
For most "Things", you don't need tons of bandwidth and the extra cost of a WiFi capable microcontroller or expensive daughterboard.
Consider a low bandwidth, low power transceiver on 915 MHz or 433/434 MHz(depending on region) that only costs a couple of bucks.
Re:What Are His Thoughts (Score:4, Insightful)
For most "Things", you don't need tons of bandwidth and the extra cost of a WiFi capable microcontroller or expensive daughterboard.
Unless the device is connecting over a home power network, you also need a display and some kind of touchpad to configure wifi. Like my WiFi printer, if I want to connect my coffeepot to my wifi it needs a means to show me a list of SSIDs it can 'see' (and/or a means to key in a hidden one), then a means to enter a password. It also needs to display connectivity state. I suppose you could put a USB port on the coffeepot and then configure it with your laptop, but that gets annoying, fast.
(Almost, but not quite, as annoying as having your coffeepot online in the first place).
Re: (Score:2)
NFC.
http://finance.yahoo.com/news/eliminate-complicated-wireless-set-tis-110000089.html [yahoo.com]
Re:What Are His Thoughts (Score:5, Insightful)
What are his thoughts on the absolute saturation of the 2.4GHz spectrum and the increasing load on the 5GHz spectrum.
The internet of things sounds nice, but the amount of devices pointlessly WiFi enabled and broadcasting in my home is having a very negative effect on the usability of the spectrum for anything of value. Does he have any plan to mitigate this growing issue? Does he care at all about it or is he solely focused on funding and an early exit with a fleeting; 'I'm rich, bitch'?
Obviously, barring a miracle in the ISM band, more stuff chattering isn't going to help; but (some) of these 'internet of things' widgets are forced to confront the problem, albeit because they are too cheap or too power constrained to just shove a full 802.11b/g/n chip in there and scream there little hearts out.
In order to accomodate severely cheap and/or battery powered devices, Contiki provides support for 802.15.4-based networks (through 6LoWPAN), which are both much less resource intensive and rather less chatty than 802.11 devices.
Arguably, the overall effect of 'internet of things' chaff on larger computers trying to use the spectrum probably depends on adoption, with three rough possible trajectories:
1. Apathy: The perceived value of the 'internet of things' is roughly nil, so aside from a few horrid proprietary wireless meter-reader systems and things, field deployments are negligible, and thus so is RF traffic.
2. Partial adoption: This is actually the worst case: Wifi is a terrible mechanism for the purpose; but unlike the dreadful stew of incompatible and often partially or wholly proprietary low-power/low-speed links, it has the distinct virtue of being built into just about all the computing devices you already own, which creates a certain incentive for people building lightbulbs and thermostats and similar stuff to shoehorn it in; because they can't be sure that you'll be able to speak any other wireless protocol unless they provide a dongle (which won't work with your smartphone, and obligates them to be in the 'helping idiots install peripherals' business, which isn't somewhere you want to be). So, multiple-years-on-a-CR2032 type devices won't happen, or will be dongle-to-device; but there will be a whole lot of mains-connected devices abusing a high-throughput, noisy, protocol to dribble tiny amounts of data back to the mothership.
3. Ubiquity: This case is worse in terms of absolute number of devices chattering away; but imagines a sufficient interest and volume of deployment that, say, it becomes expected for home routers and the like to speak some standardized 802.15.4-based protocol, so there are indeed a great many devices; but they aren't forced to speak wifi for compatibility reasons. The ISM band continues to be a mess; but at least devices get to choose a protocol roughly commensurate with their actual throughput requirements.
Internet of Crap (Score:3)
A scene from my future. (Score:1)
Joe The Garage Door Opener Repair Guy finishes up his work and wants to connect my new garage door opener to the Internet, so that I can open the garage door from work or with my state-issued iDevice or some other bullshit I don't need to live my life, because the little remote clipped to the visor in my vehicle is no longer good enough for the job for some reason because everyone is a technology fetishist these days, and don't spend one second considering whether something that can be done is something tha
"Open source" (Score:2)
Nice marketing plan. Thingsquare will provide software for all the little devices as an open-source reference design (hyped as "Open Source" everywhere on their web site), which will encourage companies to use it (and then likely close it, being BSD) for devices they manufacture. This will provide Thingsquare with a large collection of compatible devices with little effort on their part. On the other hand, it is all useless without a server to manage everything, which of course is provided by Thingsquare
The problem with the internet of things (Score:2)
That the software that manage it will be open source could mitigate some of the problems (lower odds of software backdoors pass unnoticed for much time), but a