Orientation device

[glennword]

I was thinking about designing a swerve drive, and how to tell the robot that it was facing the opposite way from which it started. I thought of a radio compass, in which a signal is transmitted from a device on the driver station, and the robot could then tell what its orientation was. One problem: is there a rule preventing me from using such a device?
Please post any ideas/tips/comments!

[RyanN]

A gyro can be used, but drift must be accounted for if used over the entire match.

A magnetometer (compass) would work as well, but you should consider high current electronics, motors, ferrous metal, overhead metal objects, rebar in concrete, etc...

I tried to use a magnetometer in 2008 Overdrive. It was the biggest headache... I then tried a magnetometer to orient an antenna tracking system for my school's UAS team, and the motors in the servos threw it off.

My recommendation is to avoid magnetometers at all cost, and try to implement a gyro.

[Alan Anderson]

If I were a robot inspector, I'd have to tell you that a radio transmitter would not be permitted. For example:

Originally Posted by 2011 <R52>

One D-Link DAP-1522 is the only permitted mechanism for communicating to and from the ROBOT during the MATCH. All signals must originate from the OPERATOR CONSOLE and/or the Field Management System, and be transmitted to the ROBOT via the official ARENA hardware. No other form of wireless communications shall be used to communicate to, from or within the ROBOT (e.g. radio modems from previous FIRST competitions and Bluetooth devices are not permitted on the ROBOT during competition).

A standard passive electronic compass might work, if you isolated it sufficiently from things like motors and power wires.

[Jon Stratis]

I wanted to add this, from the 2012 Inspection checklist:

Unauthorized Wireless Communication – no wireless communication to/from robot or operator console without FIRST permission. Laptop wireless (WiFi and Bluetooth) disabled. No radios allowed on the operator console or in the pit<R55, R84, Admin Section 4.3.1 >

[dtengineering]

There are specific, fixed points on the field (and in the ceiling above the field) that emit light. It may be possible to use a camera to track orientation.

Just don't ask me to write the code for it!

Jason

[ToddF]

Extrapolating from Jason's idea of using ambient light, there are also ambient radio waves being transmitted from fixed locations around the field. Though wireless communications are not permitted, there isn't any rule against passively detecting sources of transmitted radio waves. Just like a cell phone can triangulate it's location from cell towers, it might be possible to triangulate robot position and orientation. You could use one of your practice matches to drive around the field and make a calibrated map.

It would be a heck of an engineering challenge, though. The necessary equipment might also put you way over budget.

Todd F.

[Gdeaver]

As dtengineering suggested the lights in the ceiling would provide a frame of reference outside the robot. A magnetometer also provides a frame of reference outside the robot. Triangulation with a radio transmitter also provides this. Of the 3 the mag is the least difficult. Our team is looking at using 9 degrees of freedom. A mag, an accelerometer, and a gyro with sensor fusing algorithms. The big problem is the mag and the hard and soft iron distortions. We did some testing in September with our 2011 robot. The 2011 robot is ideally suited for a mag. there is very little ferrous metal or motors up high. Our 2012 robot has no area on it with out magnetic distortions. If we use this we will have to design the robot with it in mind. The latest generation of Gyros has improved allot. I feel that field centric robot control with out operator intervention is now possible. It does take allot of effort.

[Bryce Paputa]

you might be able to do something like an optical mouse (or multiple ones), I wouldn't be sure how though, and it might not work if the field isn't flat. you could even use something like a mechanical mouse, you could just have some omniwheels with encoders, you should be able to figure it out with 4? you might need more and it would probably take a lot of DIO channels, but i think it would be possible.