Whegs Rover Gait Tuning for Rough Terrain

ISEF Category: Robotics and Intelligent Machines

Subcategory: Robot Kinematics  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A robot can look fine on a flat floor and fail the moment the ground gets messy. That gap is where whegs rovers get interesting. You can tune how the legs move, then see which settings actually help on bumps, gravel, and carpet. That turns a cool robot into a real research project.

What Is It?

A whegs rover blends wheels and legs. Instead of rolling on round wheels, it uses legs that move in a repeating pattern. That pattern comes from a central pattern generator, or CPG, which is a control model inspired by the body rhythms that animals use for walking. You can think of it like a metronome for robot motion. Change the beat, and the robot steps differently.

Your project asks which gait settings help the rover move better on rough surfaces. Phase means how far apart the leg motions are in time. Amplitude means how far the legs swing or move. In simulation, you can search many phase and amplitude settings fast, then test the best ones on a real rover. That gives you a clean link between robot control and real motion on terrain.

Why This Is a Good Topic

This is a strong science fair topic because you can test it with clear variables and measurable outcomes. You can compare speed, slip, stability, and energy use across different terrains, then ask which gait settings work best and why. The project connects to search-and-rescue robots, planetary rovers, and machines that need to move through messy places. You can learn simulation, controls, experimental design, and data analysis in one project.

Research Questions

• How does leg phase offset affect rover speed on printed bumps?
• What is the effect of gait amplitude on slip ratio in gravel?
• Does carpet favor a different CPG phase pattern than a hard floor?
• To what extent does the best simulation gait match the best hardware gait?
• Which terrain type causes the largest drop in stability across phase-amplitude settings?
• What is the effect of gait frequency on energy use while keeping phase and amplitude fixed?
• How does obstacle spacing change the phase setting that gives the best traversal?

Basic Materials

• Whegs rover platform or similar legged rover testbed
• Laptop with Python support
• PyBullet installed on the laptop
• Arduino or other motor controller, if your rover uses one
• Rechargeable battery pack matched to the rover
• Phone camera or basic video camera for motion tracking
• Printed bump course made from cardboard, foam, or 3D prints
• Gravel tray
• Carpet sample or carpet tile
• Measuring tape
• Digital stopwatch
• Marking tape for start and finish lines
• Notebook or spreadsheet for data logging
• Flat calibration ruler for image scaling.

Advanced Materials

• Research-grade whegs rover or custom build with encoder feedback
• Motor drivers matched to the rover motors
• Microcontroller with logging support
• IMU for pitch, roll, and acceleration data
• Wheel or leg encoders
• Force sensor or current sensor for power estimates
• 3D printer for terrain modules
• Adjustable test rail or overhead camera mount
• High-speed camera, if available
• Motion capture access, if available
• Load cell for traction testing
• MATLAB or Python environment for parameter sweeps
• Dedicated workstation for batch simulation.

Software & Tools

• PyBullet: Simulates rover motion on custom terrain and lets you sweep gait parameters.
• Python: Runs parameter searches, data cleaning, and plots of speed, slip, and stability.
• ImageJ: Measures displacement and body tilt from video frames.
• Tracker: Tracks rover position in recorded video for simple motion analysis.
• Google Sheets: Organizes trials and helps you compare terrain conditions quickly.

Experiment Steps

1. Define the motion metrics you care about, such as forward speed, slip, stability, and energy use.
2. Build a small simulation model that lets you change phase and amplitude without changing the rest of the rover design.
3. Choose a terrain set that includes at least one smooth surface and several rough surfaces with different textures or obstacle shapes.
4. Plan a parameter sweep so you test a wide range of gait settings before you ever touch hardware.
5. Select the best simulation candidates, then design hardware trials that keep the terrain and starting conditions consistent.
6. Compare simulation and hardware results, then look for patterns where the model succeeds or breaks down.

Common Pitfalls

• Changing motor voltage between trials, which makes gait settings look better or worse for the wrong reason.
• Using terrain pieces that shift during runs, which changes the contact conditions from trial to trial.
• Measuring only travel distance, which can hide slips, stalls, and unstable body motion.
• Trusting one simulation setting without hardware validation, which can miss friction and compliance effects.
• Comparing results across surfaces without matching the starting pose, which makes phase settings look inconsistent.

What Makes This Competitive

A strong version of this project goes beyond, which gait is fastest. You can build a real map of performance across phase and amplitude, then compare that map across several terrain types. If you add slip, power, and body attitude data, you get a richer picture of why one gait works better than another. The best entries also test how well simulation predicts hardware, because that mismatch is a real research question.

Project Variations

• Test the same gait search on a rover with different leg stiffness or foot materials to see how contact changes the best phase setting.
• Replace the terrain set with loose soil, foam blocks, or a stair edge to compare how obstacle type shifts gait performance.
• Add an analysis layer that compares forward motion, slip, and power draw, then rank gait settings by a combined score instead of speed alone.

Learn More

• PyBullet Documentation: Read the simulator docs and examples to learn how to build custom robot and terrain models.
• MIT OpenCourseWare, Introduction to Robotics: Search MIT OpenCourseWare for robot kinematics, locomotion, and control notes.
• NASA Technical Reports Server: Search for rover mobility, legged locomotion, and terrain interaction reports.
• IEEE Xplore or arXiv: Search for review articles and papers on legged robots, wheeled-legged robots, and CPG control.
• PubMed: Search for review articles on central pattern generators to understand the biological idea behind gait rhythms.
• USGS and NOAA data portals: Use these for terrain, soil, and surface context if you want to connect rover design to outdoor conditions.