No they are not.
Vast array of hardware options (from cheapest MEMS gyros to "fancy" MEMS gyros to electrical and optical gyros - prices from a few $ and terrible performance to $100.000 and more and very good performance).
Then this needs to be backed up at least by accelerometers to vector Earths gravity whenever this is possible (in un-accelerated flight or on the ground, standing still).
The trouble with gyros is they are "relative". They do not give you an absolute view of the horizon. They simply telling you that you are rotating a bit this way and a bit that way. It's not terribly accurate so measurement errors quickly build up and your horizon goes out the window.
One solution is to carefully measure each gyro error source during manufacture and compensate during operation. That can give you several minutes of good performance even with relatively low cost devices (good enough to easily pass TSO standards with a fair margin). But the calibration is a "bitch" to say the least. This is what we do and using the accelerometers in all sorts of tricky ways to assist - this still qualifies as "unaided" as only inertial sensors are used. We have a higher grade solution as well using British Aerospace designed gyros but the principle is exactly the same - we still have errors, just a bit less.
We have a laser gyro system in development as well for space based applications but this is quite pricey. However, even this has, surprisingly, quite a few errors so nothing as changed - just less errors.
OK, that is one way of doing it - then there is another. We are doing this "other" way in our low cost iBOX for the iEFIS. This is the "aided" way. Here you use velocity information from airspeed or GPS to figure out centrifugal and other forces on your aircraft so you can subtract these from your accelerometer readings - the result, if all is perfect (which it is not) is that you can still vector earths gravity in certain maneuvers (just which ones depend on how clever you are). This way means that you do not have to rely much on gyros and only need them to measure rapid CHANGES in rates. So in effect, they only need to be good enough for a few seconds. That means you don't have to spend time on calibrating them much - almost plugin and go.
Which is better ?
Depends - no straight answer is possible, it depends on your needs.
A good, unaided system is the best solution as it can follow you through sustained aerobatics for a long time and it does not need any additional inputs.
A typical, low cost aided system will assume that you are flying "nicely" so it can use the velocity to figure out what bank angle you are at during a turn - as long as it can assume that the "ball" is in the middle. However - throw anything else at it and there are soon issues. Yes, you can do a quick roll or two but sustained aerobatics is only possible with the next version of aided system which I will describe now:
It is possible to build an aided system that uses a particular type of GPS that can use low cost gyros AND be very usable during sustained aerobatics. It's a complex system from a software point of view but not too expensive to make. Trouble is, it does need a good GPS, excellent antenna arrangements and of course a good satellite constellation. In effect this system is an extension of the normal "aided" system mentioned above but is able to measure velocity in three dimensions using the GPS so it becomes possible to fairly accurately compensate all three axis of the accelerometer (in other words, the system knows where "down" is even without following the gyro derived horizon calculations).
If you have ever flown behind one of our later EFIS systems that is not using an AHRS but is using the GPS to calculate a "horizon" you can see aiding in action. This is effectively a good version of the aiding that would be used in a 3D GPS aided AHRS - minus the gyros and minus the accelerometers. It's pretty convincing - but like any aided system, it stops working when the aiding information source stops (GPS in this case).
And that's the downside of the aided system - it needs something else to work. As long as it's working and the designers of the aiding algorithms have done a good job - it will work just as intended. If the aircraft is doing something the aiding cannot interpret correctly, aiding can be pretty bad and do the wrong thing.
The good news is, I think, that currently all manufacturers have many years experience with what they have been doing and have constantly developed and tweaked their systems. I'm pretty sure all systems available now should be able to pass TSO certification with relative ease (TSO does not call for aerobatics - the phugoid test being perhaps the most difficult to pass - and even that is easy now).
Rainier
CEO MGL Avionics
