Predict crowd congestion before it happens
CrowdSynthetic is an open‑source AI crowd safety simulator that predicts congestion before it becomes dangerous. It visualizes movement, generates real‑time heatmaps, scores zone‑level risk, and triggers automated evacuation logic. Built for concerts, festivals, temples, and stadiums, it helps organizers understand crowd behavior and prevent disasters through simulation and analytics.
CrowdSynthetic provides a configurable layout-driven simulation with real-time heatmap visualization, avatar movement, and zone-based capacity tracking. Users can model concert-style venues with stages, exits, and zones to understand crowd flow patterns before events occur.
CrowdSynthetic includes EVAC mode logic that automatically triggers evacuation based on crowd risk thresholds. Users can observe evacuation behavior, validate exit effectiveness, and review metrics logged during emergency scenarios.
The tool uses trend-based prediction to forecast saturation ('FRONT in X seconds' / 'FRONT saturated') and generates AI recommendation text. It logs capacity-based zone risks to CSV and JSON, enabling risk assessment and evacuation planning analysis.
The simulation exports detailed metrics to concert_risk_log.csv and concert_risk_log.json, providing researchers and safety planners with quantitative data on crowd dynamics, zone saturation, and risk progression for post-event analysis.