Color Sensors

[Dean Hystad]

Originally Posted by bananabots

From my looking at the board, I'm guessing it is more likely color sensors will be used to detect medicine bottle color!

That will depend a lot on the rules.

Back in "No Limits" (I think) there was a mission designed with the light sensor in mind. There were three similar field models, one a different color. The models were randomly placed in three different positions at the start of the match. To score you had to activate the odd colored model. Most of the teams I saw solved this without using any kind of color sensing.

The bad cell destruction mission from "Body Forward" had a different twist. Here the cells were randomly toggled any time the robot was not interacting with the model. I saw no way that the mission could be solved without doing some kind of color sensing. I don't know if it was the sensor requirement or the difficulty of manipulating the model, but I saw very few teams attempting the mission.

It looks like the medicine mission has a new twist. Looking at where everyone is predicting the models go I see no markings indicating an exact placement. If the order is random the mission will have a lot of variability. Will its prime location close to base or number of points be enough to entice teams to solve the mission?

The Ball model is interesting. The color of the center ball has to mean something, and that color can be affected by either team. The ability to quickly identify which ball is in the prime spot and take corrective action may be vital.

[RVDHRobotics]

Originally Posted by bananabots

From my looking at the board, I'm guessing it is more likely color sensors will be used to detect medicine bottle color!

My kids discussed this option, but felt that the orange would probably reflect more than the green enough to be able to identify the bottles. That and they don't know how to program a color sensor.

[bananabots]

Originally Posted by RVDHRobotics

My kids discussed this option, but felt that the orange would probably reflect more than the green enough to be able to identify the bottles. That and they don't know how to program a color sensor.

I would suggest you encourage the kids to try a quick experiment to see if their hunch is correct. In my limited experience with color sensors, they actually do a pretty good job at color determination and are quite easy to program.

There are even ways you can see real-time sensor outputs from within the NXT-G programming environment. So you can move the sensor around and see what color it is determining live.