Last year we used the Analog Devices Gyro sensor to help us navigate our turn in auto mode and really liked how it worked. This year we’d like to also try and use linear accelerometers as well.
Any teams out there try them before and to what success were they?
Analog Devices has some app notes on using an accelerometer to measure incline. Could be useful to know when a robot started to tip or when it was on a ramp. Note that they are also good vibration sensors you’ll need some hardware and software filtering.
We’ve been using the Analog Devices accelerometer in the off-season to see how useful it may be to help control velocity. Although we have had some positive results, there are a few draw-backs and it is not nearly so simple to use as the yaw rate sensor for measuring heading.
The drift on this sensor is much worse than Analog’s yaw rate sensor.
The sensor has high frequency response. However, it has an on-board filter which can cut out much of this content.
When you spin, if your sensor is not located at the center of mass, there is a centripetal acceleration component which provides a bias in one direction.
Although I’m sure we can ultimately find a way to over-come all of these problems, we still have a lot of engineering to go and will probably not attempt it in 2005. Or, if we do attempt, it will probably not be a critical component.