Chandni Sanariya
Blue Cohort
11/9/07
Lab 3: Line Following: Summary
Summary of Results
My program was pretty successful. It did most of what it was assigned to do. If both sensors were on white it drove forward. If one of the sensors was on black it would turn slightly in that direction until both were on white. (On black if cliff readings are less than 750; on white if greater.) The only times this was unsuccessful was when there were sharp turns, corners, or cusps. The create wasn't able to correctly follow the path in those instances. For example, the create may not have been able to turn quickly enough or it may have moved further before tackling that part of the track first.
Important Lessons Learned
It is a must to start small and simple with your program and continue by adding little by little. If you don't start little, you won't know where to look and what to fix if there's an error. Try to be as simple as possible. And make sure you understand everything you're using.
A lot of programming is trial and error. You have to see what works and what doesn't and use that in your program to make it more efficient.
Your program might not execute the way you envisioned it, so it is important to put yourself in your robots place to help program. This is why those comments are helpful because it is what you expect the function will do and what the create should be doing.
Inherent Limitations
* (1)In this scenario, the robot would bot be able to detect the right turn and then the immediate left turn as the robot follows the path. It would continue until it sensed a black line when the path finally turns enough to get in the robot's way.
- (2)In this instance, it would be difficult for a robot to choose a path on its own. It would have to be programmed to chose a certain path.
- (3)In an instance where there is a tight turn, it is a challenge for the robot to make that turn because it might pass the line after making the initial turn.
 |
If I could add on stuff to the Create, I would add on...
I think if I could add anything to the robot to make it line follow more efficiently, I would add a middle cliff sensor right in between the left front and the right front cliff sensors. By having that we could set up a condition where the robot only goes forward if the two outer ones were on white and the middle on black tape. This way, you can ensure that the robot is on a line at all times, instead of empty white space when you wouldn't have a middle cliff sensor. And the robot would only turn if either of the two outer ones were on black. This could probably minimize some of the limitations up above like the last one. I was also thinking, that if IR sensors worked the same way as cliff sensors, then we could add one that could detect the colors/light on the ground and use that to guide us the same way the cliff sensors do.
Back to Lab 3 Line Following
