Optical mouse sensor

The title says it all, we really think an optical, mouse type sensor would be very useful in autonomous programming.

Well… there aren’t really any mice in the game. What’s the point? I mean, I guess maybe you have some interesting hobby project you’d like to work on.

Hobby project like making a small robot to follow a mouse through a maze? I’ve always wanted to try that!

I dunno how useful a mouse detector would be in Toss Up though.

EDIT: A robotic mousetrap would also be useful.

It would work similar to encoders but would not be subject to errors from wheel slip. It would be of little use (or difficult to use) in Toss Up because of the bump, but could provide some pretty awesome accuracy in past an possibly future games.

I know it would not be a “Mouse” being used in the way that it would be utilized on a robot, but that description was the simplest.

They really wouldn’t be any more accurate because as soon as it lifted off the surface a little then you no longer get data, just like with wheels. Furthermore, you’d need 2 of them because otherwise you can’t sense robot rotation. They really wouldn’t provide any service the OSEs don’t already provide.

The problem you may have with autonomous isn’t accuracy, it’s prediction. You can already figure out pretty exactly how far you’ve traveled, the problem is that you need to know to slow down at the right point so you end up traveling exactly what you want to travel. The current sensors are fine for this.

Although I do admit it would be pretty cool technology to play with. But just to play with - not for any practical reasons.

O.K., I admit you busted me! It would be a toy :smiley: I am Sure James Pearman or Chris Siegert have already figured out how to interface an off the shelf mouse.

Haha sorry didn’t mean to be a killjoy! :smiley:

It would be very cool, seriously! Just not practical for a competition

Yes, it’s been done a bunch of times, here’s one.

http://hackaday.com/2012/09/15/robotic-odometry-from-an-optical-mouse/

The problem, as always, is legality. We cannot use it for competition but that doesn’t stop you from doing it for fun.

We looked at this for a Trinity Firefighting competition robot where we’re not so constrained by Vex parts but are constrained by a 12" cube. Went through the research but were a bit stumped at how to read in the values properly and ensure a good data rate.

Here are some additional resources…

Nice explaination of how the optical mouse works here:
http://botsnlinux.net/school_projects/mouse_paper_colloquium_22march11.pdf

Another one found here:
http://people.duke.edu/~dh67/files/MPS.pdf

Nice break down here of the hardware…
http://www.youblob.com/node/213

That leads to some sample code in BASIC…
http://www.parallax.com/Portals/0/Downloads/docs/prod/hardware/28560-MouseSensorKit-v1.0.pdf

Make magazine had something as well…
http://blog.makezine.com/2007/12/15/using-an-optical-mouse-for-rob/
which links here…
[http://home.roadrunner.com/~maccody/robotics/croms-1/croms-1.html

At this point the regular methods (the normal Vex sensors) were deemed good enough for the time at hand…

Maybe someone more adventerous can tackle this into the Vex system or and Arduino board.](http://home.roadrunner.com/~maccody/robotics/croms-1/croms-1.html)