UTM Logo
Agent-Based Modeling (ABM)

Flood Evacuation Simulation Logic

An educational journey through algorithmic simulation development. Explore how environmental variables, human behavior, and infrastructure affect disaster response.

Explore Versions

Technical Concepts Explored

A* Pathfinding

Agents use the A* (A-Star) algorithm to calculate the shortest path to safety. The simulation dynamically updates the "cost" of grid cells—flooded roads become impassable (infinite cost), forcing real-time re-routing.

Cellular Automata

The environment uses a grid-based logic where the state of a cell (Dry/Flooded) is determined by its elevation, the global water level, and time-based weather triggers.

Emergent Behavior

Complex crowd dynamics (like traffic jams at bridges) emerge not from a central script, but from individual agents following simple rules: "Don't walk into an occupied space" and "Move to high ground."

Simulation Evolution

Version 1.0

Terrain & Pathing

The foundation of the simulation. Focuses on generating procedural terrain and implementing the core A* pathfinding logic.

  • Random map generation
  • Basic Resident Agents
  • Simple Elevation Logic
Version 2.0

Environment & Time

Introduces a central river, a bridge bottleneck, and a simulation clock. Agents react to weather changes based on personality (Cautious vs Stubborn).

  • Day/Time Cycle
  • Weather States (Rain/Storm)
  • Agent Risk Tolerance
Version 3.0

Social Dynamics

Adds physical collision and queuing. Agents can no longer "ghost" through each other, creating realistic traffic jams at the bridge.

  • Collision Detection
  • Bridge Bottlenecks
  • Traffic Statistics
Version 4.0

Intervention

The final iteration adding user interactivity. Deploy a Search & Rescue helicopter to airlift residents trapped by the traffic or floodwaters.

  • Controllable Entity (Heli)
  • State Management (Flying/Landed)
  • Advanced UI Dashboard