Line Follower Sensitivity?

We just installed the line follower kit onto our robot and for some reason it’s really sensitive. We tested it with pure white tape on yellow (And tried blue) foam pieces. The white tape gives a reading of ~40, but the blue or yellow foam gives a reading of ~60, sometimes even 45. Is there a problem with the sensors or are they supposed to be like this?

You want the line to be high-contrast against the surface. I’ve used black electrical tape on a white melamine surface and got a large difference in the readings.

It is possible that the foam you are using is more reflective to IR than visible light, making it appear brighter to the line tracker than it does to your eye. I’d experiment with different surfaces using the On-Line Window until you find a combination that gives you good results.

Cheers,

  • Dean

Will it be like this in the competition with the gray foam pads?

The closer you get the sensors to the mat, the more precision (tighter range) you get. But even with the sensors close to the mat, the readings will depend upon the ambient light

We were getting readings of 30-600 for white and 750+ for black over a range of light settings. Those readings tightened when we lowered the sensors until they practically touched the mat (more like 30 - 50 for white, 750 - 800 for black).

Our team’s programmers have toyed with trying to follow the white line on the gray competition mats. They found that differentiating between the 2 colors was difficult and were more interested in pursuing other strategies than trying to get an accurate line follower.

I have never been to a competition, but I have used the line following sensors at home using the gray foam tiles like they use there. If you look at the sheets that came with the line following sensors, you will see that it does recommend mounting the sensors fairly low - about 1/8th of an inch I believe.

I am using mine to detect black and white tiles which are darker or lighter than the gray tiles, respectively. If I remember correctly, I think I was getting readings of aprox. 40-45 for the white tiles, aprox. 150-250 for the gray tiles, and somewhere over 900 for the black tiles. Sorry I don’t remember the exact numbers, but they didn’t vary that much in my house.

I was using the tiles in our kitchen and we have windows all along the South wall so I had plenty of light coming in. There was plenty of difference between the readings for white and gray tiles to reliably tell them apart in a program., so it should work with the white tape too.

Before I decided which color of tiles to use, I took one of the line following sensors and held it over each of the color samples they sent me to check the readings. For most colors like white, blue, red, yellow, green, etc. the readings were around 40-45 at the proper distance.

Hope this helps you decide what you want to do. Don’t give up on the line following sensors if you really want to use them.

I have found that placing the line follower sensors underneath the main body of your robot that it blocks out the ambient light. Contrast is still the main key here. Colors other than black or white will work with these sensors, but will keep giving you similar results as you have been getting. Red for instance is about a 50% grey when you change it to grayscale in Photoshop. Even though it may stand out on a lighter surface, the contrast is not there.

Fun Fact: In the original Psycho movie by Alfred Hitchcock, they had to actually change the blood in the shower scene to another substance because the greyish fluid did not look like blood in black and white. They ended up using Hershey’s Chocolate Syrup.

Getting the sensors as close to the floor is key. The blue tape / red tape on the gray backgrounds are not as good as white tape but they can be seen.

A bright light may throw the sensors off, but room light has not been a problem for me.