The inhabitants of urban environments live their daily lives in the same space, and for their mobility share the same infrastructure. The question of how to enhance mobility while at the same time reducing congestion, accidents and pollution is a common challenge to all major cities in Europe.
By focussing on fulfilling the mobility demand of the Berlin population in the upcoming 2020s it is essential to add new opportunies of urban mobility to the given static network. Jelbi hubs are closing the gap between public transport and individual mobility possibilites without requiring a major intervention in the existing infrastructure.
To provide as much convenience as possible and efficient possibilities of quick adaptions of the demand, new locations for Jelbi hubs need to be chosen wisely.
What it does
We are combining individual routing requests to the BVG (demand) with open data sources of the given public transport network of Berlin (supply) to identify the current demand for placing new Jelbi hubs.
The Berlin public transport network of the BVG and the S-Bahn is characerized by its given stops, routes and frequencies of lines plus time tables and fares. Based on this network, our goal is to develop an improved decision support of such a strategic planning problem for new Jelbi hubs.
How I built it
- Modelling the demand
- Cleaning, parsing and aggregating the provided data (~300Gb of raw csv data) of the routing requests to the BVG
- Merging the data of requests for routes with the geospatial data of the BVG and S-Bahn stops
- Add Open data sources (Open street map, ...)
- Visualizing the geospatial data with geopandas and geoplot
- Build animations to show the changing request density over daytime
- Modelling the supply
- Analyzing the structure of the public transport network (VBB-Netz Berlin/Brandenburg)
- By using a Voronoi-diagram we identified the nearest stop for a given location (commuter belt)
- Deriving locations with high and low densitiy of public transport stops by graphic visualization
Challenges we ran into
- Cleaning the data of the provided requests to the BVG App (Encoding, bad data, ...)
- Joining geospatial data of locations and requests (different files)
Accomplishments that I'm proud of
- Visualization of the data (none of us is experienced with frontend)
- Working with new unknown team mates
What I learned
- Visualizing geospatial data with QGIS
What's next for ImproveJelbi
- Adding more specific open source data, such as
- demographic data
- information of big events (concerts, sport events, demonstrations, ...)
- traffic data
- long-distance traffic
- dedicated locations such as the airports, Olympia Stadium, Waldbühne, Tempodrom, Wuhlheide, ...)
- Regarding time dependent demand (seasonal effects, weekend vs week, morning vs noon, ...) of Jelbi hubs and find a solution for adapting the supply of a Jelbi hub dynamically
- The long term objective tries to find a compromise between quality and cost.