Rush Hour: Crowd Dynamics Simulation

Rush Hour is an advanced behavioral simulation designed to model the chaos of human crowds. Transforming the client’s original arcade game concept into a serious planning tool, we use Unity’s DOTS architecture to simulate tens of thousands of individual AI agents navigation complex environments like malls, stadiums, and train stations.

Safety in Numbers

Preventing crowd crush and bottlenecks requires predictive modeling. Rush Hour allows safety officers to test evacuation plans virtually, identifying dangerous pinch points before a facility even opens its doors.

Challenge & Solution

The Challenge: Managing large crowds involves complex variables. Traditional “flow charts” cannot account for human irrationality—panic, grouping behavior, or distraction. The client needed a way to visualize how 50,000 people would exit a stadium during an emergency, or how a Black Friday sale would impact mall corridors.

The Solution: We engineered Rush Hour, a massive-scale agent simulation.

Individual Behavior

Each agent in the simulation has a unique “Personality Profile” (e.g., Aggressive, Slow, Group-Oriented). Using navigation meshes and local avoidance AI, they naturally form lanes, clusters, and bottlenecks, realistically mimicking real-world crowd dynamics.

Metrics:

Flow rate (people per minute)
Density heatmaps
Evacuation time analysis

Final Result

Rush Hour has become an essential tool for safety certification in large venues.

50k Agent Sim

Successfully simulated 50,000 unique agents in real-time at 60fps, providing granular data on mass crowd movements.

30% Better Flow

Venues optimized using Rush Hour saw a 30% improvement in ingress/egress speeds during sold-out events.

Safety Certified

The simulation data was accepted by local fire marshals as valid evidential support for new evacuation capacity ratings.

Rush Hour demonstrates that understanding chaos is the first step to controlling it.

Summarize with AI:

We Give Unparalleled Flexibility