I left my boyfriend and dog back in Florida when I came back to ITP in August. Sometimes the distance between yourself and the people you care about can be challenging. Certainly though there are many ways to stay connected from afar.
There are a lot of communication tools and a surprising number of them act as surveillance tools too. SMS/iMessage, FaceTime, Facebook, Find My Friends, WiFi Scales, WeMo, etc. I started looking at this with Dropcam, exploring public/private spaces and the way the perception of being watched influences behavior. The original purpose of installing the camera was to supervise Georgia the dog when we first adopted her. When I moved back to New York, I felt that it was unreasonable to expect my boyfriend to give me 24/7 uninterrupted video steams into his apartment without offering the same. So that's how that came to be.
Georgia the dog wears Whistle, a Bluetooth Smart Ready / WiFi enabled device that provides analytics about her behavior. (And a future version will be the first device to use the SIGFOX network to provide GPS tracking information). Essentially it's a FitBit for your dog. The Whistle team has built a nice iOS app but doesn't provide any data on the web, and no API.
[side note, apparently now Flickr is forcing this iframe embed]
It's a shame that Whistle doesn't offer an API. The data and trends are pretty interesting and I've noticed that Georgia's activity level correlates with how my boyfriend's week is going. I can surmise the mood of the day of my boyfriend based on the activity of the dog.
This is before he's home from work so it just includes a morning walk.
Trends paint the picture, Georgia is designed for 60 minutes of activity per day.
Time including when I still lived in Florida.
Could I improve the mood of my boyfriend by convincing him to walk Georgia more often?
Whistle API-ish + NodeJS + Twilio
I wasn't the only person interested in getting raw data from Whistle. Someone else reverse engineered their APIs and released a nifty node wrapper. Once I got a key we were ready to go. I'm parsing the JSON responses to get the current activity time and the next connection time for the device. It usually connects to WiFi every hour.
I have a pretty crude proof of concept working. This is an example of some of my test messages:
This message style is way too mechanical. But the data is getting where it needs to go.
So far he's been less than enthused by Georgia's communication. But I did get Georgia to 46 minutes of activity tonight, she only had 27 minutes yesterday and 18 the day before.
He appears to be taking the messaging in good spirits, he even added the Twilio number as a contact for the dog. In the next few days it should become clear if it's actually increasing the activity level of both him and Georgia.