I've been reading up on gyroscope data collection and it sounds as though gyroscope readings tend to "drift" over long periods. Would there be a good way of tracking or correcting for this drift?
You need some other way to measure absolute attitude in order to correct for the long-term drift of your (rate) gyros so that you can use them for meaningful short-term measurements.
Some systems use accelerometers to measure the gravity vector to establish roll and pitch references (but this requires that there are no other accelerations involved) and magnetometers to establish a yaw reference.
Other systems use multiple (two or three) GPS receivers to do interferometry on the GPS signals in order to establish an absolute attitude reference. This is harder to do, but can be done even while on the move (accelerating).
The most sophisticated systems combine all four kinds of information (rate gyro, acceleration, magnetometer and GPS, typically using a Kalman filter) in order to provide the most robust solution.