Pillbug Paths for Better Robot Exploration

ISEF Category: Robotics and Intelligent Machines

Subcategory: Other  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A tiny bug can beat a robot at finding its way through a maze of open space. Pillbugs do not follow a neat map, but their wandering can hide a smart search strategy. If you track those paths, you can turn animal behavior into a robot rule set. That gives you a rare mix of biology, coding, and robotics.

What Is It?

This project studies how pillbugs move through a controlled arena and asks whether a robot can copy that movement pattern. Pillbugs are small land-dwelling crustaceans that often explore by changing direction, hugging edges, and reacting to local cues. You track their paths in a 3D-printed arena with OpenCV, which is a computer vision tool that helps you turn video into coordinates.

Think of it like learning a travel style from someone who never uses a map. You are not trying to make the robot act like a bug for fun. You are testing whether a biologically inspired exploration policy can cover space faster, more evenly, or with fewer repeats than a standard robot search pattern.

Why This Is a Good Topic

This is a strong science fair topic because you can measure real behavior, turn it into data, and test a clear engineering claim. The project connects animal navigation, computer vision, and robot path planning. You can ask whether nature gives you a better exploration rule than a simple random walk or grid sweep. You also get room to learn video tracking, behavior coding, and efficiency metrics without needing a giant lab setup for the analysis itself.

Research Questions

• How does the pillbug colony’s average path efficiency change across arena shapes?
• What is the effect of edge density on pillbug coverage of the arena?
• Does a robot that copies pillbug turning behavior cover more unique area than a random-walk robot?
• To what extent does obstacle placement change the repeat path rate of pillbugs?
• Which movement metric, turning angle, stop frequency, or wall-following time, best predicts colony-wide coverage efficiency?
• How does the robot’s coverage efficiency compare when it uses a pillbug-derived policy versus a fixed search pattern?

Basic Materials

• 3D-printed or laser-cut arena with removable walls.
• Pillbugs from a safe classroom or field source, with proper care container.
• Smartphone or webcam for overhead video capture.
• Tripod or fixed camera mount.
• Uniform backlight or lightbox to reduce shadows.
• Nonreflective arena surface.
• OpenCV-compatible computer.
• Notebook or spreadsheet for behavior coding.
• Metric ruler or grid sheet for arena calibration.

Advanced Materials

• High-resolution overhead camera.
• Microcontroller or robot base with differential drive.
• Wheel encoders for path logging.
• Custom 3D-printed arena inserts with variable obstacle layouts.
• Environmental sensors for humidity and temperature logging.
• Tracking software pipeline built in Python with OpenCV.
• ROS or similar robot control software.
• Statistical analysis environment for path comparison and model testing.

Software & Tools

• OpenCV: Tracks pillbug positions from overhead video and extracts path coordinates.
• Python: Cleans tracking data, calculates coverage metrics, and runs the exploration model.
• ImageJ: Checks video frames, measures arena dimensions, and helps verify tracking quality.
• R: Runs statistical tests and compares robot performance across conditions.
• GeoGebra: Helps sketch arena layouts and test geometry ideas before you build them.

Experiment Steps

1. Define the exact behavior question you want to test, then choose one coverage metric that matches it.
2. Design the arena so the camera can see every path clearly and the robot can later use the same space.
3. Plan your tracking pipeline, including how you will detect the pillbugs, smooth the coordinates, and handle missing frames.
4. Decide which pillbug movement features you will convert into a robot rule, such as turning bias, wall following, or stop frequency.
5. Build comparison trials that include at least one baseline robot policy, so you can judge whether the biology-based rule helps.
6. Set up your analysis plan before collecting the final data, including how you will compare coverage, repeats, and path diversity.

Common Pitfalls

• Tracking pillbugs under uneven light, which creates false movement spikes in OpenCV.
• Using an arena surface with too much glare or texture, which makes the bugs hard to detect reliably.
• Mixing individual behavior with colony behavior, which can blur the policy you are trying to model.
• Letting the robot and the pillbugs operate in different arena conditions, which makes the comparison unfair.
• Choosing only one coverage metric, which can hide tradeoffs between speed, repeat visits, and area uniformity.

What Makes This Competitive

A stronger project goes beyond a simple path trace. You would build a clean tracking pipeline, test more than one exploration metric, and compare the bug-based robot against strong baselines. You could also ask whether the policy still works when the arena shape changes or obstacles move. That kind of careful comparison turns a cool demo into real research.

Project Variations

• Test whether pillbug paths change when the arena has more wall contact points.
• Compare a pillbug-derived robot policy against a pure random walk and a lawnmower-style sweep.
• Analyze whether humidity or light level changes the colony’s coverage efficiency and path repetition.

Learn More

• OpenCV documentation: Learn the core video tracking tools and find examples in the official OpenCV docs.
• NIH PubMed: Search for review articles on insect locomotion, collective movement, and bio-inspired robotics.
• NOAA Open Data and educational pages: Review motion tracking, image analysis, and pattern detection methods used in environmental science.
• NASA Image Analysis resources: Explore how researchers extract coordinates and measurements from video and images.
• MIT OpenCourseWare, Introduction to Robotics: Review robot motion planning and path efficiency concepts through free course materials.