Active Matter Swarms in Maze Geometry

ISEF Category: Physics and Astronomy

Subcategory: Biological Physics  ·  Difficulty: Advanced  ·  Setup: School Lab  ·  Time: Full Year

The Hook

A swarm can look smart even when every robot follows simple rules. That is the weird power of active matter. You can test how walls and turns change that group behavior, first on a laptop, then with a real robot swarm.

What Is It?

Active matter means many self-moving particles that keep pushing on their own. Think of each bristlebot like a tiny puck with a motor. Each one follows simple motion rules, but the group can form bands, clusters, or swirling loops.

A Vicsek model is a simple simulation where each particle tries to align its direction with nearby particles, plus a little noise. An active-Brownian-particle model adds random wandering, like how dust jitters in water, but with self-propulsion. You can use either model to ask how crowd motion changes when the space is open, narrow, or maze-like.

Confinement geometry matters because walls can trap motion, redirect flow, and create loops. A printed maze is like a set of traffic barriers for tiny movers. If you compare simulation results with a bristlebot swarm in a 3D-printed arena, you can test whether the same geometry shapes both the model and the real system in similar ways.

Why This Is a Good Topic

This is a strong science fair topic because you can change one thing, the wall geometry, and measure clear outputs like swirl count, cluster size, and direction order. It connects to robotics, swarm behavior, and biological physics, so the idea has real-world meaning. You can start with a simple simulation, then add physical validation with cheap robots, which gives you both theory and experiment experience.

Research Questions

• How does maze corridor width change the order of motion in a Vicsek swarm?
• What is the effect of wall angle on the chance that particles form a stable swirl?
• Does higher noise make a swarm less likely to get trapped in loops inside confinement?
• To what extent do circular arenas and maze arenas produce different cluster sizes?
• Which maze shape produces the strongest long-lived circulation in an active-Brownian-particle model?
• How does particle density affect whether a swarm breaks into one large cluster or several small ones?

Basic Materials

• Laptop with Julia installed.
• Free 2-D plotting software or notebook environment.
• Basic access to a 3D printer or maker space.
• 3D-printed arena pieces or maze inserts.
• $30 bristlebot parts or small vibration bots.
• Small batteries or coin cells matched to the bots.
• Ruler or calipers for arena dimensions.
• Smartphone camera for overhead video.
• Masking tape for marking the test area.
• White paper or matte board for background contrast.

Advanced Materials

• Access to a 3D printer for custom arena geometries.
• High-frame-rate overhead camera for motion tracking.
• ImageJ or Python-based tracking setup for particle paths.
• Motorized vibration platform for a more controlled active-particle analog.
• Precision scale for matching robot masses.
• Laser-cut or machined inserts for repeatable confinement shapes.
• Access to a physics or robotics lab for calibration runs.
• Optional small IMU or sensor modules for robot motion checks.

Software & Tools

• Julia: Runs the Vicsek or active-Brownian-particle simulation and lets you test geometry changes quickly.
• ImageJ: Tracks bristlebot paths and measures swirl patterns from overhead video.
• Python: Cleans trajectory data and calculates order parameters, cluster sizes, and circulation metrics.
• GNU Octave: Offers free matrix and plotting tools if you want a second analysis path.
• Jupyter Notebook: Keeps your simulation notes, plots, and comparison checks in one place.

Experiment Steps

1. Define the exact motion rule you want to test, then decide whether Vicsek, active Brownian motion, or both will be your main model.
2. Design a small set of confinement geometries, such as open space, corridor, corner, and maze loop, so you can compare them fairly.
3. Choose one or two outputs that capture collective behavior, such as alignment, circulation, clustering, or time spent near walls.
4. Build a simulation first, then tune it until it produces sensible baseline behavior in an open arena before you add maze boundaries.
5. Plan a physical arena that matches the simulation dimensions closely enough for a real-world comparison with bristlebot motion.
6. Set up a tracking method that converts video into trajectories, then compare simulation and swarm results with the same metrics.

Common Pitfalls

• Using a maze that is too complex, which makes it hard to tell whether the walls or the particle rules caused the behavior.
• Letting the simulated particle density differ from the bristlebot density, which breaks the comparison between model and experiment.
• Measuring only one video run per arena, which makes random noise look like a real geometry effect.
• Tracking robots with changing camera angle or lighting, which distorts paths and swirl counts.
• Comparing a simulation with ideal point particles to real bristlebots without accounting for collisions, wall slip, or turning bias.

What Makes This Competitive

A stronger project will not just say that mazes change swarming. It will explain which geometric features matter, and it will back that up with clean metrics. You can raise the level by comparing more than one model, testing several arena shapes, and using statistics that separate real effects from random motion. A tight link between simulation and physical swarm data also makes the work feel much more original.

Project Variations

• Test how a single sharp corner versus a rounded corner changes swirl formation in the same swarm model.
• Compare a simulated particle swarm in printed mazes with a real bristlebot swarm in mirror-image arenas.
• Analyze whether wall friction or wall shape matters more by keeping the maze geometry fixed and changing the surface finish.

Learn More

• MIT OpenCourseWare, Statistical Physics and Active Matter lectures: Search MIT OpenCourseWare for courses on statistical physics, soft matter, or active matter.
• PubMed: Search review articles on active matter, collective motion, and swarm robotics for background reading.
• arXiv: Look for preprints on Vicsek models, active Brownian particles, and confinement effects.
• NIH PubMed Central: Find free full-text papers on collective motion and biological physics.
• Reviews of Modern Physics: Search for review articles on active matter and flocking to get a deep overview.