spa-l3-paris-emergency-routing
Map (reference ← swipe → agent)
0:00
Need the dispatch coverage model rebuilt before Thursday's briefing. The eight historical emergency-call locations are in `incidents.csv` (columns: incident_id, latitude, longitude, label). Pull the driveable road network and all hospitals from OpenStreetMap via Overpass for the area bbox 48.83,2.30 to 48.88,2.38. I need three deliverables: the shortest driving route from each incident to its closest hospital, a distance matrix of each incident against its three nearest hospitals, and a 15-minute drive-time isochrone around every hospital to visualise coverage gaps. Assume 30 km/h for edges without a maxspeed tag; use posted maxspeed otherwise.
Everything goes into a single GPKG called `emergency_routing.gpkg` with all layers in the official metric coordinate system for Paris and exactly these four layer names. The `incidents` layer has Point (or MultiPoint) geometry and an `incident_id` column. The `closest_hospital` layer has LineString (or MultiLineString) geometry representing the route, with columns `incident_id`, `hospital_name`, and `network_distance_m` (numeric, metres). One row per incident. The `distance_matrix` layer is tabular (geometry may be empty or null) with columns `incident_id`, `hospital_name`, `rank` (integer 1 through 3, ascending by `network_distance_m` within each incident), and `network_distance_m`. The `isochrones_15min` layer has MultiPolygon geometry (Polygon also accepted) with columns `hospital_name` and `travel_time_min`. One row per hospital.
0:00
System info
{
"adapter": "openrouter",
"model": "google/gemma-4-26b-a4b-it",
"base_url": "https://openrouter.ai/api/v1"
}
0:01
Bash
ls -F
ls -F
incidents.csv
0:02
Read
incidents.csv
1 incident_id,latitude,longitude,label
2 INC_001,48.8566,2.3522,near Notre-Dame
3 INC_002,48.8620,2.3360,near Louvre
4 INC_003,48.8462,2.3464,Quartier Latin
5 INC_004,48.8700,2.3431,near Gare du Nord area
6 INC_005,48.8530,2.3325,Luxembourg Garden area
7 INC_006,48.8600,2.3700,Bastille area
8 INC_007,48.8450,2.3100,near Montparnasse
9 INC_008,48.8750,2.3600,near Republique
0:17
Write
/work/check_incidents.py
content · 6 lines · py
import pandas as pd
df = pd.read_csv('incidents.csv')
print(df.head())
print(df.info())
Wrote /work/check_incidents.py (89 chars).
0:18
thought<tool_call|>