Safety-Assessment

This project is based on ride-data gathered by the SimRa mobile application. The aim is to map different street segments to infrastructure types, that could be extracted from openstreetmap. After mapping, scores regarding the popularity, safety and mixed-popularity (a score gained from popularity and safety) are calculated for each infrastructure-type and saved in a postgres database. After computation the results can be visualized with the provided notebook. An overview and explanation of the resulting database can be found here.

Local development setup:

1. Follow the instructions for setting up the postgres database
2. Import the postgres dump (consists of SimRa rides mapped to openstreetmap street segments) provided here* by running the code below:

gunzip -c simra.sql.gz | psql simra

3. Install the code dependencies:

pip install -r src/requirements.txt 

4. Add the project to your PYTHONPATH.
5. Run the code:

python src/app/main.py

To specify the areas scores should be calculated for, change areas.json. Note: Score calculation will only work for areas included in the SimRa dataset. To make sure that the areas you want are included check the dashboard.

*For security reasons the download is protected by a password. To gain access message @tobiasengelbrecht or @vDawgg.

Startup using docker-compose:

1. Unzip the database and place it in the db folder
2. Make sure the areas included in areas.json fit your needs (running more areas than necessary will take a long time)
3. Run the following to start the database and a python container with jupyter notebook

docker-compose up -d

3. Access the notebook at localhost:8888. Once inside the notebook run the first cell to start the import process. On first startup it will take some time until you are able to run the importer, as the db is being imported. Make sure to download the results and graphics for later use. If you want to look at different areas, adapt areas.json and run docker-comopse up --rebuild safety_assessment.

Note: This is intended as a version for easier setup. If you only want to look at the results in the notebook this is fine. Otherwise, we recommend using the setup above.

Environment Variables:

Key	Value
POSTGRES_HOST	'localhost' ('simra_db' when running in docker)

The .env file should be in the Safety-Assessment directory.

Project Structure:

> doc                   // Further documentation, already linked to here
> notebooks             // notebook for visualizing and analyzing the results
> src                   // source-code, Dockerfile and requirements
    > app               // areas.json and main.py
        > calculation   // score calculation
        > city-results  // results from included cities in csv-format
        > data          // database and overpass functions