ThingSpeak: Building My Own Twitter (for Things)

Over the past several months I have been working on software to allow “things” to form social networks and send status updates via the Internet. At first glance this may sound very impractical. Hopefully, in a few years this will make more sense as better applications come out. Remember how ridiculous my toaster sounded three years ago? I gave that thing a voice and since has been on TV and more people are interested in it than my Twitnot not saying ter status updates. More proof? At CES 2011 there were at least 10 appliances that could send Twitter status updates. I am not saying that I created them, but I am that they didn’t know about My  Toaster and it’s 600+ followers on Twitter. With this project, I wanted to take it a few steps further and build something from the ground up that’s focused on collecting enormous amounts of data from everyday objects, allowing devices to interact with each other, and building applications to present some meaning. The ThingSpeak project is finally ready to go and open to anyone that wants to start building applications.

My Problem

One problem with the Internet of Things is the concept of “the killer app” – the app that defines a new industry. The internet connected refrigerator is our poster child. Recently, Ryan Rusnak connected a mini fridge to the iPhone via ioBridge, added a motor controlled beer selector, and strapped on an air cannon to fire beers to his couch from 25 feet away – now that’s a killer app. So, my problem was trying to find an application that highlights key features of ThingSpeak and why it’s different.

My App

What I come up with is the idea of “my_house”. “my_house” is a collection of “rooms” that all hold computers, appliances, and sensors aka things. I recently installed some light sensors in “my_room” that push light levels to the ThingSpeak API. I did this to remind me when to turn my lights on so I don’t work in the dark when I get plugged in. And an interesting side benefit was that now I can detect whether or not someone is in the room based on the light level. I will admit this could have been done other ways, but sometimes you solve problems with what you have lying around. “my_house” already keeps track of my location using Google Latitude to control my thermostat.

Eureka moment…

Now I can tell when my mom goes into my room when I am out on the town, most likely Sheetz!!!

Below is a screenshot of the demo app and how I aggregate the data collected by my light sensors. Here are a few things I want to point out:

  • “my_house” is sending the status updates via a collection of networked sensors
  • The status updates were generated by the light sensor device itself
  • The timestamps were recorded by the ThingSpeak API
  • If you click the chart icon you get what the actual light level was
  • The app works in real-time and you can check it out here

"Twitter of Things" ThingSpeak Demo App

The Technical Details

The light sensor uses a Netduino Plus that connects to my home network over Ethernet. This device uses Microsoft .NET Micro Framework and I wrote an application that interfaces with the ThingSpeak API. I also wrote a tutorial over at the ThingSpeak community site on how to use the Netduino Plus for those that want to get started with it fast.

The front-end application is written using only HTML, CSS, and jQuery. I have the fully documented source code attached and you can also see the app live in your web browser.

Light Sensor Netduino Plus

Project Files:

Leave a Reply

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