When it's time to study or socialize in a public space, there are many options that a college campus affords. But what if you know that you need a good spot specifically distraction-free studying, or want to find a place where people are already hanging out?
Roomie is a sensor network that monitors the status of rooms on campus and sends this data to a publicly available map so that you can check in on where to go to do what you want to do.
I thought it would be interesting to solve this problem with a wireless sensor network because each node can be made to be fairly low-cost, low-resolution, and high efficacy. Having a simple light and sound sensor that's network-enabled allows for a fair bit of useful data collection.
The light and sound data is interpreted as follows:
Lights on, loud room: social
Lights on, quiet room: study
Lights off, quiet room: empty
Lights off, loud room: creepy.
The collated data is displayed with text and color clues to increase speed of reading. Below are some screenshots of one sample room's status as displayed based on the sensor data.
A full realization of this network would use wifi-enabled sensor devices, running on low power (ideally solar power, if possible), distributed all over indoors public spaces on campus. The data from these would feed into a central server, which would then publicize the data to a dedicated website and also publish it via an API, so that app developers could integrate it as they wish.
Processing and Arduino code below:
I had quite a few unexpected stupid technical challenges, including more than an hour and a half of inability to get an Arduino to talk to Processing—still somewhat unresolved. It appears that for some reason wireless serial communication interferes with an amplified electret's output.
You can upload files of up to 20MB using this form.
