Lab 8

Summary

This lab was putting together all the previous labs done until this point to allow our robots to perform stunts. As we were allowed to be truly creative in this lab, it allowed me to stretch my imagination for what is possible with the RC car robot.

Procedure

Controlled Stunts

For this lab, I chose to implement the first stunt option, which involves driving the robot as fast as possible at a wall, then right before it hits the wall, flipping backwards and driving back from where it came from.

I chose this task for a couple of reasons. First, this task seemed to be in line with the task I chose to do in Lab 6, which meant less work than the other option. Second, this task seemed to be the most practical option considering how I could default back to a completely open-loop solution if things went south (and they did… somewhat). Most importantly, this was a cool task that would make for some interesting clips, as you will see.

To time the flip for the robot, I used a PID controller, similar to my solution in Lab 6. However, because the robot was moving at the highest speed (this was required for the robot to flip successfully), I learned quickly that the ToF sensor values were too delayed. My solution to this problem was by increasing the 500mm threshold in order to account for the high velocity and the processing delay, which worked out pretty well in the end.

void stunt()  // Continuously called in the main loop
{
  int waypoint = 500;  // This value was tuned as I ran tests
  if (pid(tx_tof_f_value, waypoint) > waypoint) {
    drive(FORWARD, 255);
  } else {
    drive(BACKWARD, 255);
    delay(1000);
    drive(STOP, 0);
    delay(500);
    drive(BACKWARD, 255);
    delay(1000);
    drive(STOP, 0);
    stunt_running = false;  // Stops calling stunt() in the main loop
  }

Here is a graph of the PWM values of the motors:

Here is a video of the robot performing the stunt. Even though the robot did not technically cross over the start line at the end, it certainly would have if it did not change angles after the flip, which is sort of out of my control unfortunately:

Open-Loop, Repeatable Stunts

Now that the required stunt was out of the way, I was free to try other, exciting stunts. I wanted to be very ambitious at first and tried to get the robot to draw a nautilus shell-like spiral shape, but I soon realized that this was a risky stunt for the robot to repeatedly perform, as is evident by these videos:

(Needless to say, I am worried about the structural integrity of the robot after these runs. Yikes)

So, I settled for an easier, safer stunt: spinning really fast in place. This is a lot more boring, but I figured it was the best option without spending too much extra time on stunts.

Here is the (very simple) control sequence:

for (int i = 0; i < 10; i++) {
  drive(RIGHT, 200);
  delay(50);
}
drive(ASTOP, 0);

Bloopers

Please enjoy these bloopers at the expense of my (and the robot’s) painful moments:

Conclusion

Overall, this lab was a fun one! It was rewarding to put all the knowledge from the previous labs into practice.

References

• Lab handout

Back to main