#ReactivaMadrid Hackathon

Context

In May 2020, as Spain began lifting COVID-19 lockdown restrictions, Madrid City Hall ran the #ReactivaMadrid hackathon — challenging participants to propose data-driven solutions for safely reopening the city’s public spaces. The challenge: how do you allow people to return to parks, markets, and pedestrian areas without creating dangerous crowd concentrations?

Role

Team member and technical lead for the data pipeline and geospatial analysis. The team was multidisciplinary — I handled the Python modelling stack while teammates contributed urban planning domain knowledge and the final presentation.

Approach

We built a crowd flow modelling tool that combined geospatial data with mobility signals:

• ArcGIS for spatial data ingestion and visualisation of Madrid’s public space network.
• GeoPandas for polygon-level analysis of crowd capacity estimates per zone.
• Mobility modelling: used publicly available foot traffic data patterns to project crowd density at different times of day, with configurable capacity caps per zone.
• Risk scoring: each zone received a risk score based on crowd density vs. available surface area, with a simple red/amber/green output for city managers.
• FastAPI prototype endpoint serving zone risk scores for potential integration with a city dashboard.

The project was completed in 48 hours with a small team.

Outcome

• 2nd place out of all participating teams.
• Solution presented to Madrid City Hall representatives and urban planning technical staff.
• Demonstrated that open geospatial data combined with simple modelling could provide actionable crowd safety guidance with minimal infrastructure.