Bristle-Bot Swarm Motion and Directed Drift

ISEF Category: Robotics and Intelligent Machines

Subcategory: Biomechanics  ·  Difficulty: Intermediate  ·  Setup: School Lab  ·  Time: 1 to 2 Months

The Hook

A tiny robot can look lost one second and purposeful the next. That switch can happen when you change one vibration setting or a leg angle by a little bit. Your project asks where that line is. You will turn messy motion into data, then test a model that predicts when the swarm stops wandering and starts moving with purpose.

What Is It?

Bristle-bots are simple robots with angled bristles or legs that move when vibration shakes them. Think of them like a toothbrush head turned into a tiny crawler. The bristles grab the surface unevenly, so vibration becomes forward motion instead of just noise.

A swarm adds another layer. When many bots move together, each one bumps, turns, and reacts to its neighbors. That can create random walk behavior, which looks like a drunk person walking with no clear direction, or directed motion, which means the group drifts with a preferred path. Your project looks for the bifurcation, the point where small changes in vibration frequency or leg angle flip the whole system from one motion mode to the other.

A Langevin equation model helps describe that motion. In plain language, it treats movement as a mix of push, drag, and randomness. If your data match the model, you can explain not just what the bots do, but why they do it.

Why This Is a Good Topic

This makes a strong science fair topic because you can change one setting at a time and measure the result with video tracking. The phenomenon connects to robotics, swarm behavior, and biomechanics-style motion analysis. You can collect real data at school, quantify paths, and compare them with a physics model. That gives you room to ask a clear question, test it carefully, and build a result that feels like research, not a demo.

Research Questions

• How does vibration frequency change the fraction of bristle-bot paths that show directed motion??
• How does leg angle affect the average speed of a single bristle-bot and a swarm??
• What is the effect of swarm size on the transition from random walk to directed motion??
• To what extent does surface texture change the bifurcation point for directed motion??
• Which leg-angle and frequency pairs produce the highest path straightness in video-tracked trials??
• Does the Langevin-equation model predict the turning-rate distribution across different vibration settings??

Basic Materials

• Bristle-bot components or prebuilt bristle-bots.
• Eccentric vibration motor and battery pack.
• Assorted bristles, wire, or angled plastic legs.
• Testing surface panels with different textures.
• Smartphone or camera for overhead video.
• Tripod or fixed stand for the camera.
• Ruler or grid mat for calibration.
• Tape for marking the test area.
• Stopwatch or timer app.
• Digital scale for measuring bot mass.

Advanced Materials

• Access to a vibration table or controlled motor drive.
• High-speed camera or a camera with stable frame rate.
• Calibrated force sensor or accelerometer for vibration characterization.
• Laser-cut or 3D-printed bot bodies with adjustable leg angles.
• Multiple surface samples with measured roughness.
• Optical tracking markers or printed fiducials.
• Computer with data analysis software.
• Spare motors, batteries, and fasteners for repeated trials.
• Safety enclosure for repeated vibration tests.
• Reference calibration object for video scaling.

Software & Tools

• OpenCV: Tracks bot position frame by frame from video and extracts paths, speed, and turning angle.
• Python: Processes tracking data, fits models, and runs statistics.
• ImageJ: Checks video calibration, frame quality, and motion blur before tracking.
• GeoGebra: Helps plot curves and inspect where motion shifts from random to directed.
• LibreOffice Calc: Organizes trial data and makes quick graphs for screening patterns.

Experiment Steps

1. Define the motion metric you will use, such as path straightness, turning rate, or net displacement.
2. Choose the single main variable you will sweep first, then lock the others as controls.
3. Plan a video setup that keeps scale, lighting, and camera angle fixed across trials.
4. Build a tracking workflow that turns each path into numeric features you can compare.
5. Fit a simple motion model, then test whether the data change sharply near a threshold.
6. Design a second comparison, such as a different surface or swarm size, to check whether the threshold shifts.

Common Pitfalls

• Changing camera position between trials, which makes path length and speed measurements inconsistent.
• Mixing bots with different leg angles in the same run, which hides the effect of the variable you meant to test.
• Using a surface that slowly shifts or vibrates, which adds fake drift to the paths.
• Tracking every bot with the same model settings, which can fail when bots overlap or blur.
• Treating random scatter as a real transition, which can happen if you do not compare multiple trials and controls.

What Makes This Competitive

A stronger project will not just report that the bots move differently. It will pinpoint a threshold, compare several surfaces or swarm sizes, and test whether the same model still works across those cases. Better entries also quantify uncertainty and show whether the transition is sharp, gradual, or different for each setup. If you add a careful model fit and a clean control comparison, your project starts to look like real research.

Project Variations

• Test how different surface roughness levels shift the random-to-directed motion threshold.
• Compare single-bot motion with swarm motion to see when group effects change the bifurcation.
• Replace leg-angle sweeps with motor frequency sweeps and analyze whether the model parameters stay stable.

Learn More

• OpenCourseWare from MIT: Search for introductory dynamics, vibration, and nonlinear systems lectures that help you model motion.
• PubMed: Search review articles on locomotion, gait, and swarm behavior to compare your framing with biological motion studies.
• Journal of Experimental Biology: Search for papers on locomotion mechanics and animal-inspired motion models.
• NASA Image and Video Library: Find examples of motion tracking, visualization, and experimental setup ideas.
• OpenCV Documentation: Read the official tutorials on video capture, tracking, and contour detection.
• Python Scientific Libraries Documentation: Use NumPy, SciPy, and Matplotlib docs for analysis, fitting, and plotting.