EcoBici CDMX Trip Data Scraper

Extract historical trip records and live station data from EcoBici — Mexico City's Lyft-operated public bike-share network with 480+ stations and over 10 million trips per year.

What this actor does

• Downloads and stream-parses historical monthly trip CSVs going back to 2010 (97 MB+ per monthly file)
• Fetches live GBFS station data — station names, coordinates, dock capacity, and real-time service status
• Enriches trip records with origin and destination station details (lat/lng, capacity, operational status)
• Respects memory constraints via streaming CSV parsing — files are never buffered in full

Input

Field	Type	Default	Description
mode	string	both	What to fetch: trips (historical CSVs), stations (live GBFS), or both
yearFrom	integer	current year	Earliest year to include in historical trip data
yearTo	integer	current year	Latest year to include in historical trip data
maxItems	integer	unlimited	Cap on total output records

Output

Each record contains:

Field	Description
trip_id	Row-level identifier (YYYY-MM-<index>)
bike_id	Bike identifier from the source CSV
user_gender	Rider gender (M/F)
user_age	Rider age in years
origin_station_id	Numeric station ID (origin)
destination_station_id	Numeric station ID (destination)
origin_station_name	Station name from GBFS
destination_station_name	Station name from GBFS
origin_lat / origin_lng	Station coordinates
destination_lat / destination_lng	Station coordinates
depart_at	Departure datetime (ISO 8601)
arrive_at	Arrival datetime (ISO 8601)
duration_seconds	Trip duration in seconds
trip_type	docked (standard docked bike)
month	Source month (YYYY-MM)
year	Source year
source_csv_url	URL of the CSV file
station_capacity	Total dock slots at origin station
station_status	IN_SERVICE or NOT_IN_SERVICE
source_url	EcoBici open data page URL
scraped_at	ISO 8601 scrape timestamp

Use cases

• Urban mobility research and transportation policy analysis
• Last-mile connectivity studies (Roma, Condesa, Polanco, Centro Historico)
• Station-level demand forecasting and rebalancing optimization
• Academic research on bike-share systems in Latin America
• Housing density and real estate proximity signals

Notes

• Monthly CSV files for recent years are approximately 97 MB each. Runs that request multiple years will take proportionally longer.
• Station enrichment relies on the live GBFS feed at the time of each run. Decommissioned stations may not appear in the GBFS but will still appear in historical CSVs.
• The actor uses streaming CSV parsing with no in-memory buffering — memory usage stays well below 1 GB even for large historical downloads.