Internet-of-things Pan/Tilt mounted camera using a raspberry pi

  • Linux


    Helen and I have been working off and on (mostly off) on this project since September. The project is an internet-of-things pan/tilt camera.

    The idea is that it can be setup at one location (raspberry pi and pan/tilt camera) and monitored/controlled from another location using Minecraft or a web sockets enabled app running in a webpage.

    The underlying transport mechanism is a mosquito MQTT server

    To this point it has been tested and works both from a webpage and from Minecraft.

    The 3D printed parts can be found on thingiverse.
    The code for the project is at github.


    This project is not complete … and it has no estimated completion date; the documentation is incomplete, the code is pretty rough and not all of it has been uploaded to Github (notably the Minecraft plugin).

    Ideally, the camera will change to a Raspberry Pi camera (this has already been tested) and it will get a weather-proof enclosure.

  • When you change to the pi camera, are you going to put the whole pi into the pan and tilt rig to make weather proofing the connections easier or are you planning to have the camera external? Just curious because the carrier holding the phone looks like it would easily hold the whole pi once done (and a battery pack or a simple cable route from the bottom etc)

  • Linux

    Good question … to this point the setup has only ever been used inside so not a lot of thought has gone into it yet … we used it in Vancouver when the kids demo’d at a TedX talk … it was successful but fairly finicky as it is still running off a breadboard.

    If you mount the whole setup in the carrier a lot of problems would disappear w.r.t. wired connections … except for the connection to the base motor … which would require some thought.

    Also, depending on the servo settings the servos can jerk a fair bit and the ‘jerk’ would be magnified the more weight there is in the carrier.

  • What about gearing down and just using a brush motor and stop contacts, or a stepper motor? Both would smooth it out and totally eliminate the cost. The stepper would add a fair bit of cost, a brushless with some printed gears and stop contacts wouldn’t change the price much at all and depending on your gear box it would be liquid smooth no matter the weight.

  • Linux

    Both excellent suggestions …

    My initial thought is to move from using a cell phone camera to using the raspberry pi camera module … the weight reduction will be significant … then I will have a better idea of how the servos handle the weight … I am trying to maintain a pretty simple build in the hopes that school kids can easily replicate it.

  • Thanks for posting Chris, very interesting project.