Specifically with the small satellite and cubesat people, there is an rather 'old' idea around: Passive attitude stabilization based on permanent magnets.
The concept is rather simple. A magnet, which is mounted on a satellite, will always attempt to align its magnetic field with Earth's magnetic field. This is intriguing, because it does not need computers, power, moving parts or just any form of control.
For me, there are two critical weak spots in this concept. First, if it works, it does stabilize a satellite only on a single axis. Second, Earth's magnetic field is not a clean dipole - with respect to both, its intensity and the orientation of magnetic field lines. But even if you treat it as a simple dipole, the relative rotation of the magnetic field lines along e.g. an polar orbit does not occur in one geometric plane or with a constant angular velocity. For me, intuitively, this whole thing should lead to a rather strange tumbling behaviour of such a satellite.
How is it supposed to work? Are there studies of this concept based on attitude data from satellites actually in space? If yes, what was found?
Results from detailed computer simulations are interesting for me, too, but I clearly prefer results based on flight data. There are a fair number of projects, which intend to use magnets or used them (while the satellite failed), so I find it rather hard to find statements, which go beyond 'it will work'.