Home Automation Via Twitter


We’ve seen tons of stuff hooked up to send updates to twitter when something happens. That’s what we expected when we first read this tip from [Matt]. We were pleasantly surprised to see he’s actually doing the opposite. He has rigged several items to an IoBridge module and he can control them by sending tweets to his account. In the video you can see him sound a buzzer and change a display on an LCD. Its nice to see twitter being used as part of the control as opposed to updating us every time someone flushes.

25 thoughts on “Home Automation Via Twitter

  1. Why twitter? Just because you needed an asynchronous communication mechanism? And whats the practical utility apart from the cool hack?

    To display a message I could use jabber protocol to hook something up with gtalk, and have the whole laptop screen flash the message. Initiated from my iphone using some gtalk client. Gets me the same without any hardware

    The light sensing was good, but again, why twitter?

  2. jeetu,

    No particular reason why I used Twitter besides the fact that I needed a reason to play with the API.

    I have now added Facebook support as well.

    I agree with you, there is no practical use for this. I was just tired of seeing all the devices that tweet useless information.

  3. Twitter has the system in place to be used by various devices, and a good uptime record, so that’s why using twitter here seems appropriate to me, and it’s a nice twist on things, good job.
    I think you should define a control sequence and not just use common words like light LCD though, set a standard, it’ll become a new RFC and be known as the MATT sequence maybe :)

    As for a use, I think there are uses to think up aplenty surely.

  4. Its not the twitter interface that bothers me, I used to have my computer accept commands via twitter, before I figured out how to use cpanel to pipe email messages to php.

    What bothers me, personally, is the use of the ioBridge. And not the fact that your paying 100 dollars for something that can only do 4 different things at a time without reconfiguring it,t though that can be expanded with PIC’s or AVR’s, its the fact that you are adding another point of failure to your system.
    Anytime you make a net enabled system, you automatically have 2 points of failure, your home, if the power goes out, or your net connection dies. And however your accessing it. Your pretty much not going to get around that.
    With this, instead of interfacing directly to your computer, you go through a third party.
    If they were to go out of business, your SOL, if their dataserver goes down, your SOL.

    Also, your letting a third party have complete access to your system. Granted, its a company, so it most likely wont be abused.

    But I would much rather put together an old p2 or p3 box, load up linux, run it as a server, and issue commands and whatnot directly to it. Which is actually going to be the base unit for my home automation project once I can get another PIC programmer.

  5. Not very long ago I built a small email notifier that is controlled by pyython scripts and reacts to tweets sent to a certain account(mine). It would be the easiest thing in the world to change the source code and have the device light up for example a lamp, air conditioner and/or ventilation, when being told to do so, via twitter.
    Also the other way round is easy too: there’s a project called “OBDEV data logger” I think, which sends sensor values to your PC, and uses almost the same hardware/wiring schematic (an attiny45 and six resistors and two diodes, thats all).
    And there you have it already – all done with attiny45, some python scripts and twitter API. No proprietary I/O interface card necessary, full automation of whatever, via twitter.

    Side note, you can configure twitter accounts to be not visible to the public, anbd if you tell noone about this account you can well use it for your own automation steering purposes. You could even implement a password check when fetching tweets via twitterAPI so each tweet would need to include this authentication before it could effect something. Just my 2 cents..

  6. novastar,

    I do agree that having a 3rd party server in the mix is not ideal. If something happens to them, my system is completely hosed.

    But you can’t argue with ease of use, I completed the whole project in half a day.

  7. If I understand the (tiny) diagram on the project site it looks like the iobridge is connected to a pc that actively monitors twitter for updates.

    What about just installing apache & rss/blogging software directly on the pc instead? This would likely cut down on the lag time we see in the video.

  8. Matt,
    You are right, The way you did it is excellent if you dont want to spend a lot of time, and have the extra cash.

    And granted, if I had the extra cash, id probably buy an ioBridge to play around with too.

    Me, personally, being cash strapped, cant go your rout, but putting a old computer together from the contents of my closet and making up circuit boards to control what i need it to is the only way I can go.

  9. Seems a bit odd to complain twitter might go down, it’s far more likely your own site goes down since twitter is of such a size and so much used by so many people (many of whom have pull) that is has plenty of redundancy, as for privacy, well you don’t have to use plain text and it’s not that sensitive to control something at home really, and as is said you can set twitter to be private.
    Email is in fact less secure because countless countries store email traffic and the NSA and such monitor international email.
    Incidentally about secretive twitter: people can find you on twitter by searching on email address, beware.
    I do agree though that it’s nicer and more hack-like and perhaps has less lag (does twitter ever lag?) if you use more basic communication with your own setup.
    And I hear there was some trouble in china and the first thing they did is close down twitter, so if you are in foreign places you might be blocked from access if you use twitter as a vehicle and can’t find a proxy.

  10. I think after sensing the turd the tweet should smack the guy who didn’t flush until he flushes it himself. There is no reason to automate a task as easy as flushing.

  11. I use twitter as much as anyone I know, but it’s really just another protocol and frankly one that’s not so well suited for sending commands for HA. Essentially, we could take the same technology that’s been around for years, adapt it to twitter and now it gets headlines “OMG, it has teh twitters!”

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.