Line Following and Figure Eight
Right Hand Wall Following
Coming Soon
Full Robot Integration
The objective of this milestone is to make the robot successfully follow the line and traverse a grid in figure eight.
1. We were considering putting three line sensors on the robot for increased accuracy, but due to the limited amounts of analog pins on the Arduino, we ended up deciding to use two line-sensors, one on each side of the robot. Based on the input from the line sensor, the Arduino will decide whether to go straight or make a turn.
2. We started the milestone by testing the line sensors and seeing how sensitive they were. We quickly realized that the line sensor needed to be very close to the ground in order for it to read significant enough difference between black and white.
3. We then added parts to our robot to mount the line sensor and replaced the wheels with a more appropriate size. It took a lot of trial and error in order to get the line sensor to rest well and in the proper position.
4. Then we start to make the robot follow a line. To follow a line, we arranged the line sensors to be on either side of the line. This involves sensing when both sensors are above black sections. If either sensor is on the white line, the robot will readjust itself so that the sensors will again straddle the white line. This worked well, and our robot followed the line. It was even able to readjust itself when it started to go off course. The video below shows the robot following a straight line.
5. After we got the robot to follow a line, we started to work on traversing a figure eight. However, we faced a lot of issues.
6. We realized that the primary issue was with the ball bearing on the back of the robot. Because the ball bearing was unstable, the distance from the ground of the line sensors was inconsistent. This made it difficult for the robot to readjust itself. There were also times when the ball bearing was dragging behind the robot. This made it hard to turn and move efficiently.
7. We then rebuilt the robot with new wheels and a better-fitted ball bearing. This change has significantly improved the stability of sensor readings.
8. In order to get our robot to form a figure 8, we had to revisit the logic in our code. We watched the robot carefully and changed the code incrementally to reduce the amount of under and over-turning at corners. We also recalibrated the line sensors to be more sensitive to the white lines vs. the dark floor.
9. We were able to reduce many of the problems in the code by slowing the robot down and taking the turns more slowly. This allowed the robot to stay on the lines and not get lost within squares on the grid. It also gave the robot some time to readjust itself and center itself on the line after each turn.
10. We also implemented a counter in the code so that the robot knew where in the figure-8 is was. This reduced errors where the robot would turn the wrong direction. Below is the code and a video showing the robot traversing in a figure 8.