Testing

Functionality Assessment

This section documents the testing and evaluation of the PiCar-X robot's capabilities and performance.

Results shown are out of 10 trials per task.

Code Version 90° Corner Turn 3 ft Straight Line 3 ft Consecutive Curves
V1 0 / 10 2 / 10 0 / 10
V5 1 / 10 4 / 10 0 / 10
V10 3 / 10 10 / 10 4 / 10
V13 10 / 10 10 / 10 10 / 10
V15 4 / 10 10 / 10 10 / 10

Challenges

This section outlines the key challenges encountered during development and the solutions implemented.

V0
  • Connection issues: College Wi-Fi does not allow device-to-device connections, so we had to use a phone hotspot, which had limited bandwidth.
  • Hardware issues: There were no monitors that accepted HDMI, so we had to purchase an HDMI cable and a DVI-HDMI adapter.
  • Understanding the robot's file structure: There are many files in the robot's memory, and documentation on these files is incomplete.
  • Previous groups' code: Was not successful at most challenges, including following a straight line.
V1-5
  • Calibrating the robot: Aligning the wheels.
  • Track deficiencies: Lines were hand-drawn, not dark enough, and too thin.
  • Robot behavior: Due to track deficiencies, the robot constantly thought it was off the line and kept driving forward out of line.
V6-10
  • Weight distribution: Wheels cannot be turned all the way over because then the outer wheel will not be on the ground and will start sliding due to skewed weight distribution toward the back.
  • Adjusting robot's behavior: Was difficult, so we added constants to control behavior more precisely, including speed during straight, gentle, and sharp turns, reverse movements, and the angle for adjustments (gentle and sharp). We also implemented different modes that switch dynamically.
V11-13
  • Larger track: As we attempted bigger tracks, inconsistencies arose due to varying lighting and material quality (we used printer paper and Sharpie to create the track).
V14-15
  • Corner detection: The code succeeded with intersections but struggled to distinguish corners well, resulting in lower success rates at simple corners.

Videos

Watch videos demonstrating the robot's performance during various tests and scenarios.

Corner Performance

Demonstration of the robot navigating a corner using the line-following algorithm.

Full Lap

Full track lap showing the robot's overall line-following behavior and stability.