Mat Problems?

[SkipMorrow]

Our robot normally runs very straight. Freakishly straight. But lately we have been noticing a couple of spots on the mat where our robot will almost always do a short jog to the left of about one cm. The spots are very specific and the sidestep happens with near 100% regularity. Everywhere else on the mat, the robot drives perfectly. We have tried cleaning the wheels many times. The mat seems clean, but there are some wear marks in the areas. It sure doesn't feel greasy or slippery or tacky in those specific areas. I am at a loss. I am hesitant to try cleaning the mat with anything, but I'd like to hear if any of you guys have any ideas.

To be clear, it is a very specific spot. We can program in a move tank block, 10 revolutions, and aim the robot to pass over the spot. No matter how far away we are at the start, it will shift at that same spot.

[philso]

What is the table surface like under that spot? Could there be some irregularity in the table surface that is causing this?

Can you post a photo of your robot? My sons had an issue like this in the Body Forward game. We found that there was a spot on the table where it was slightly convex (1/8" over 12-15"). When the robot (with attachment) drove over this slight depression, it would turn to the left or right.

[gt0163c]

What is the table surface like under that spot? Could there be some irregularity in the table surface that is causing this?

Excellent thought. One way to test this would be to rotate the mat 180 degrees and see if the robot has the same issues in the same spots. The team could also try driving the robot on a smooth floor and seeing if there are similar issues. Seems like a great engineering challenge to troubleshoot this. The team can document their efforts and talk about them in their Robot Design Executive Summary/judging.

For cleaning the mat, I believe there used to be instructions to vacuum the mat regularly to remove dust. My team has also used a damp cloth (wet with just water) to clean the mat of dust and snew that accumulated over time.

[SkipMorrow]

I love the idea about rotating the mat! I was looking at it very carefully and I think I see an indention about the size of my hand. It's smooth, but it is enough to feel it.

The robot is full wall square robot. Axles are supported on the outside. Small plastic wheels in the back for casters.

One other thing I thought of was that our robot does have four contact points -- two drive wheels and two small plastic wheels for casters. Since there are four contact points, it is possible that there could be times when one of the drive wheels may not be in contact with the table. So we are also thinking about moving to a single, centered caster.

Of course, today the robot behaved just fine. No problems at all. We are going to watch it closely, and if we start seeing anything weird again, then we will troubleshoot some more.

[philso]

You should ask your team:
- How many points of contact with the table will your robot (2 driving wheels + 2 casters) really have on a real world table that is not perfectly flat?-
- Assuming that most of the weight is over the driving wheels, which of the wheels and casters would contact the table as your robot traverses a field that is not perfectly flat?
- What is the effect on the robot path of the caster that contacts the table?
- What is the effect on the robot path if only one caster contacts the table in one area and only the other caster contacts the table in a different area?
- What is the effect on the robot path of varying the position of the casters?
- What is the best number and location(s) for the caster(s) in terms of making the robot move on a consistent path (straight or a constant and slight curve to one side)?