I was surprised that the owner said he needed to get a Strava premium account in order to be able to auto-upload the hamster's activities. Not 100% sure why.
If you look at any neuroscience paper, mice will run 10k a day or more on a wheel overnight. I thought the axis labels were wrong when I first saw that type of study. It’s quite common.
Project:
> The main parts are an ESP32 attached to a hall sensor, which I positioned close to the wheel. Using a magnet attached to the back of the wheel, this tells the ESP32 when a full rotation has completed. It gives a timestamp (and total distance) to each magnet-pass, which is stored on the network. A python script (on my laptop) then retrieves this data in the morning, turns it into a .FIT file, and uploads it to Strava through the Strava API (with a randomly chosen run title, taken from a mostly AI-generated list of options, as I needed >100). In terms of manual additions, I had to add the photo manually, and Mollie had to do the running.
Neat! Not to toot my own horn too much, but I put together a similar project for my cat's exercise wheel. Hackaday featured it a while back: https://hackaday.com/2023/04/02/spy-on-your-cat-to-make-sure...
His exercise routine is somewhat variable but he usually goes for a short jog a few dozen times a day. Most often at 5 am, for whatever reason.
Since the article I also built and hooked up a treat dispenser for extra motivation.
> Most often at 5 am, for whatever reason.
Poo time?
Have you seen evidence he understands that running causes treats to be dispensed and is running more as a result?
This is amazing and I love it.
I was surprised that the owner said he needed to get a Strava premium account in order to be able to auto-upload the hamster's activities. Not 100% sure why.
He might be able to use https://github.com/dlenski/stravacli (written by me) or the library it's based on (https://github.com/stravalib/stravalib) to avoid this requirement and save a few bucks.
If you look at any neuroscience paper, mice will run 10k a day or more on a wheel overnight. I thought the axis labels were wrong when I first saw that type of study. It’s quite common.
Dupe of link to original post on reddit: https://news.ycombinator.com/item?id=49216926
Mollie the Hamster, on Strava: https://strava.app.link/FMubyPjMa5b
Project: > The main parts are an ESP32 attached to a hall sensor, which I positioned close to the wheel. Using a magnet attached to the back of the wheel, this tells the ESP32 when a full rotation has completed. It gives a timestamp (and total distance) to each magnet-pass, which is stored on the network. A python script (on my laptop) then retrieves this data in the morning, turns it into a .FIT file, and uploads it to Strava through the Strava API (with a randomly chosen run title, taken from a mostly AI-generated list of options, as I needed >100). In terms of manual additions, I had to add the photo manually, and Mollie had to do the running.
More here: https://www.reddit.com/r/SideProject/comments/1v9uyq0/i_buil...