Shanghai OSRM Travel Time Visualization

This project generates a high-resolution travel time heatmap from a given origin point in Shanghai using a locally deployed OSRM (Open Source Routing Machine) backend.
It builds a uniform grid of latitude-longitude coordinates, queries OSRM for estimated driving times, filters out water regions, and visualizes the results on an OpenStreetMap base map.

Click HERE to go to the project.

⚙️ Parameters

At the top of runpic.py, you can adjust the following parameters to control the experiment:

Parameter	Description	Default
OSRM_URL	URL of the local OSRM server	"http://localhost:5000"
PROFILE	OSRM routing profile (driving, walking, cycling)	"driving"
ORIGIN_LAT	Latitude of the origin point (e.g., SJTU dormitory)	31.198606
ORIGIN_LON	Longitude of the origin point	121.432725
LAT_MIN, LAT_MAX	Latitude bounds for the analysis region	30.7, 31.35
LON_MIN, LON_MAX	Longitude bounds for the analysis region	121.18, 121.86
STEP	Grid step in degrees (smaller = denser grid, slower runtime)	0.001
output_csv	Path to save raw travel-time data	"osrm_times.csv"
output_png	Path to save the final visualization	"osrm_times_map_shanghai.png"

🧭 How to Use

1. Prepare Environment

Install required dependencies (Python ≥3.10):

conda create -n osrm python=3.10
conda activate osrm
pip install -r requirements.txt

2. Start Local OSRM Server

Download Shanghai OSM data and build the routing graph:

wget https://download.geofabrik.de/asia/china/shanghai-latest.osm.pbf  
docker pull osrm/osrm-backend
docker run -t -i -p 5000:5000 -v $PWD:/data osrm/osrm-backend osrm-extract -p /opt/car.lua /data/shanghai-latest.osm.pbf
docker run -t -i -p 5000:5000 -v $PWD:/data osrm/osrm-backend osrm-partition /data/shanghai-latest.osrm
docker run -t -i -p 5000:5000 -v $PWD:/data osrm/osrm-backend osrm-customize /data/shanghai-latest.osrm
docker run -t -i -p 5000:5000 -v $PWD:/data osrm/osrm-backend osrm-routed --algorithm mld /data/shanghai-latest.osrm

Leave the OSRM server running on port 5000.

3. Run the Script

In another terminal:

python runpic.py

The script will:

1. Generate a uniform geographic grid across Shanghai.
2. Query OSRM in batches for travel times from the origin point.
3. Remove grid points falling inside Shanghai’s rivers and coastal waters.
4. Render and save a colored travel-time map.

4. Outputs

• Map image:
  osrm_times_map_shanghai.png — A heatmap of travel times:
  • Blue → fast / close
  • Yellow → medium
  • Red → slow / distant
    (Capped at 3600 s; all values above appear as red)

5. Notes

• The base map tiles are fetched from OpenStreetMap Mapnik.
• OSMnx is used to retrieve Shanghai’s administrative boundary and water polygons.
• You can adjust cmap, vmin, vmax, and alpha to tune visualization style.
• Increasing grid resolution (STEP) will produce finer results but may increase runtime.