Keeping our streets clean
Sweden is ranked as one of the most sustainable countries in the world. You can enjoy their garbage-free streets and breathe the fresh air around you. Now, this is not achieved without effort. People have a certain mindset which promotes an eco-friendly lifestyle. Sweden has a trend called plogging, which is an act of running and picking up litter (Swedish: plocka upp).
Our clients wanted to help this movement to keep our everyday environments free of pollution. From their idea and our efforts, an app called “WePlog” was born. It's a Geographic Information System (GIS) that tracks collective user activities to help them collaborate in cleaning the city effectively and easily.
The mobile application shows a colored map of the user's current location, where the colors help indicate which street segments have been already picked clean (green) and which would need more work (red). These colors are provided by taking into account sessions from previous users and parameters per street. When starting a session, a plogger can activate the app like you would on Strava, and it will record the activity in the background.
Users can also post notes and pictures that are being sent to the city administrators. This way we seek to improve communication between the community and local governments. We also introduced gamification, ploggers can gain points by cleaning streets and doing more sessions, these points can later be traded for rewards and recognition.
We also created a management tool for local administrators. This admin dashboard is used to have an overview of the current situation, and the overall cleanliness of their city. They can also check some statistics such as cleanliness score, volume of litter that was gathered, number of users and sessions over time. Using this information city admins can change the parameters of certain area's.
The challenge in this project lies in scoring streets or street segments based on raw longitude / latitude coordinates that are coming in from the mobile app. To tackle this challenge we first imported OpenStreetMap data to be able to go down to the street segment level. We used a PostgreSQL database powered by PostGIS to handle the GeoSpatial data.
We created a scoring algorithm that for each raw data point, snaps these coordinates to one of the imported street segments. Then, for each segment, it analyses these snapped locations from different users together with segment specific parameters to eventually give each segment a cleanliness score, that is used to color the map.
Some of these parameters, such as time intervals (some streets need to be cleaned more often than others) can be set manually using the admin dashboard. However we plan to optimise these parameters automatically based on user feedback on their routes using supervised machine learning techniques in the future.
Because of some great efforts by our clients so spread the word, after a few months WePog has hundreds of active users and everyday people are going out to clean their communities. It's been great to see it grow and we look forward to continue to support their journey.