Multi-Robot SLAM Under Wi-Fi Jitter

ISEF Category: Robotics and Intelligent Machines

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

The Hook

Three small rovers can look coordinated in a simulator and still get messy in real life. One weak network hop can make their maps drift apart. That gap between perfect code and real hardware is exactly what you can measure. If you like robots, maps, and data, this project gives you a real systems problem to attack.

What Is It?

This project asks a simple question with a hard answer. When 3 rovers map the same small area, how much does Wi-Fi jitter change the map they build together? Jitter means the delay between messages keeps changing. In a robot team, that delay can scramble timing, which can mess up pose estimates, map merging, and the final SLAM result.

Think of SLAM like a group drawing a floor plan while walking through a room blindfolded. Each rover makes its own guess about where it is. Then the team combines those guesses into one map. In a simulator, messages arrive on time, so the team looks neat and stable. On real Wi-Fi, packets arrive late or out of order, so the rovers can disagree about where walls, corners, and obstacles belong.

Why This Is a Good Topic

This is a strong science fair topic because you can change one factor, network jitter, and measure its effect on a clear output, map consistency. It connects to real problems in warehouse robots, search-and-rescue robots, and warehouse fleets where wireless delays can break coordination. You can learn systems thinking, experimental design, and data analysis without needing a brand-new robot. The topic also gives you a clean simulator-versus-reality comparison, which judges like because it shows careful engineering, not just a demo.

Research Questions

• How does increased Wi-Fi jitter change the overlap between maps built by 3 rovers?
• What is the effect of packet delay variation on pose estimate drift in multi-robot SLAM?
• Does map consistency stay higher in simulation than in real Wi-Fi conditions for the same rover paths?
• To what extent does message loss affect loop closure success in a 3-robot mapping task?
• Which network condition, delay, jitter, or packet loss, most strongly predicts SLAM map disagreement?
• How does robot spacing during cooperative mapping change sensitivity to Wi-Fi jitter?

Basic Materials

• 3 hobby rovers with wheel encoders and onboard computers or microcontrollers
• Wi-Fi access point with configurable traffic control or network shaping support
• Laptop for ROS 2 monitoring and data logging
• ROS 2 installed on the rover computers and laptop
• Simple indoor test area with taped boundaries for a 3 × 3 m map
• Printed markers or landmarks for repeatable navigation cues
• Measuring tape for layout checks
• Charged batteries and spare battery packs
• Notebook or spreadsheet for run logs
• Smartphone or camera for documenting runs.

Advanced Materials

• 3 differential-drive rovers with 2D lidar or depth sensing
• Raspberry Pi or similar onboard computers with ROS 2
• Managed Wi-Fi router or network emulator for controlled jitter injection
• Motion capture system or overhead camera tracking for ground truth
• Calibration targets or fiducial markers for map alignment
• Dedicated laptop with Linux for ROS 2 bag review and analysis
• External SSD or large storage for repeated data capture
• Network analyzer or packet capture tool
• High-precision measuring tools for arena layout
• Safety barriers and spare parts for repeated trials.

Software & Tools

• ROS 2: Manages robot communication, logging, and distributed messaging for the rover team.
• Gazebo: Simulates rover motion and sensor data for the idealized baseline.
• rviz2: Visualizes robot poses, maps, and message timing during each trial.
• Python: Cleans log files, computes map similarity, and runs statistics.
• ImageJ: Measures map overlays if you export maps as images for comparison.

Experiment Steps

1. Define the one network variable you will change first, such as timing variation, while keeping the rover route and arena fixed.
2. Choose one map quality metric, such as map overlap, pose drift, or map-to-map consistency, so every run produces a number you can compare.
3. Build a simulator baseline that uses the same rover motion plan and sensor assumptions as the physical setup.
4. Plan a repeatable way to introduce network stress, then match each stressed run with a calm control run.
5. Decide how you will align maps from different runs before scoring them, so small coordinate shifts do not fake a bad result.
6. Set up your analysis plan before collecting data, including the graph type, summary statistics, and rule for rejecting unusable trials.

Common Pitfalls

• Changing the rover path between trials, which makes network effects impossible to separate from motion effects.
• Comparing raw maps without alignment, which makes two correct maps look different just because they start in different places.
• Measuring only one run per condition, which hides how noisy multi-robot SLAM can be.
• Ignoring sensor timing drift between robots, which can look like a Wi-Fi problem when the real issue is clock mismatch.
• Using a simulator that does not match the real sensor setup, which makes the Sim2Real comparison meaningless.

What Makes This Competitive

A strong version of this project does more than say Wi-Fi hurts robot maps. It quantifies how much, under which kind of delay pattern, and for which map metric. You can stand out by comparing at least 2 analysis methods, using aligned ground-truth scoring, and testing whether one robot coordination strategy fails less than another. A more competitive entry also explains why the simulator misses the real effect, then backs that claim with data.

Project Variations

• Test how packet loss, instead of jitter, changes multi-robot map consistency in the same arena.
• Compare ROS 2 DDS settings to see whether one communication profile reduces map disagreement under bad Wi-Fi.
• Swap lidar for camera-based SLAM and measure whether visual mapping is more or less sensitive to network delay.

Learn More

• ROS 2 Documentation: Search the official ROS 2 docs for DDS, topics, timing, and bag recording basics.
• Gazebo Documentation: Find the official simulation guides for setting up robot worlds and sensor models.
• PubMed: Search for review articles on SLAM, multi-robot coordination, and networked robotics.
• IEEE Xplore: Search for peer-reviewed papers on multi-robot SLAM, communication delay, and Sim2Real testing.
• NASA Open-Source Software and Robotics Resources: Explore public robotics and autonomy materials, then search for rover localization and mapping examples.
• MIT OpenCourseWare: Find free robotics and control course materials that explain state estimation, sensors, and feedback systems.