Sample log point
As avid hikers and strongly conservation minded, we often find ourselves bothered by the trash and litter left behind on hiking trails. As data driven tech minded folk, we are also frustrated by the lack of a single place to log locations of trash found on trails. Given the remoteness and vastness of wilderness, it is expensive to have concerted efforts to collect data which in turn enables better decision making and resource allocation. So, the goal was to enable enthusiastic hikers like us to log data about trash we find on the trails. To lower the barrier of entry to giving data, we built both a smartwatch and a mobile application with simplicity of use front and center.
What it does
At a bare minimum cleantrails is a mobile and smartwatch app that logs GPS coordinates and timestamp when someone says they found a piece of trash and sends it to a server. The mobile app version also enables more enthusiastic users to submit pictures of trash too. At the backend, this data is saved and a holistic view is presented to decision makers. The app also uses computer vision to automatically classify the type of trash from the picture.
How we built it
This was built for iPhone and Apple watch using Swift. The devices send location data to a django server that is run on google's cloud infrastructure (app engine and google storage). The google's vision API is used to classify images received.
Challenges we ran into
Biggest challenge was version issues for iphone and smartwatch and have them communicate with each other outside of the emulator.
Accomplishments that we're proud of
We were able to get the end to end proof of concept with key components fully functional
What we learned
We had never built iphone or smartwatch apps before which was a complete first
What's next for cleantrails
Built production quality code and open source the project for people to download and start submitting data