PyBullet Swarm Sim

Fast, physics-accurate drone swarm simulation

Star

Battle Mode

Battle Mode is a competitive, multi-team simulation environment featuring kinetic elimination. Two distinct swarm algorithms (e.g., Flocking vs. PSO) are pitched against each other in a combat arena. Drones engage in physics-based combat where faster, aggressive drones eliminate slower ones upon high-velocity collision.

Battle Mode Overview

How to Run Battle Mode

You can launch a headless combat simulation directly from the command line by specifying the algorithms and team sizes:

python -m swarm_sim.battle.runner --algo-alpha flocking --algo-bravo pso --drones-alpha 10 --drones-bravo 10 --duration 20

Alternatively, for the best visual experience, use the interactive Web Dashboard to configure, monitor, and analyze battles seamlessly.

Battle Workflow

The dashboard provides a complete, end-to-end combat visualization and analytics pipeline.

1. Team Configuration

Team Configuration

Select the underlying behavioral algorithms for Team Alpha (Red) and Team Bravo (Blue). You can precisely balance the combat scenario by adjusting the drone count per team, the maximum simulation duration, and the unique job identifier for tracking.

2. Live Swarm Battle

Live Swarm Battle

Watch the battle unfold dynamically. The physics engine tracks real-time velocities and collisions. When drones collide, the one carrying lower kinetic energy is instantly eliminated and disabled. A live scoreboard and real-time kill-feed stream track every interaction.

3. Battle Results & Analytics

Battle Results

Once the dust settles, the dashboard presents comprehensive post-battle combat analytics. You can evaluate the final survival counts, K/D (Kill/Death) ratios, and structural team health to conclusively determine which swarm intelligence strategy thrives in adversarial environments.