Could airliners fetch data like AoA and speed from an INS?

Question

If I understood what an inertial navigation system (INS) is, it should be able to calculate speed and orientation of the aircraft.

If this is possible, why do airliners solely rely on sensors to measure speed and angle of attack (AoA)?

Answer 1

An INS gives your speed & orientation in space, but without reference to what the airmass that you're flying in is doing.

The Pitot tube and AOA sensors give speed and pitch orientation in relation to the local airmass, only.

An indicated airspeed of 60 knots, measured by the Pitot tube, may correspond to a stationary aircraft pointed into 60 knots of wind, or an aircraft moving 30 knots over the ground into a 30 knot headwind, or 60 knots of groundspeed in still air, or some other case. The wings fly (and stall) relative to the air, not the ground, so measurements of air data are vital.

Answer 2

The question mentions Speed and Orientation which are broad terms that could include:

Group1:
(Air Data sensor + computation)
IAS - Indicated Air Speed
CAS - Calibrated Air Speed
EAS - EquivalentAir Speed.
TAS - True Air Speed
AoA - Angle of Attack
(HDG(M) - Magnetic Heading)^^

Group2a :
(Inertial/Rate sensor + computation)
GS - Ground Speed
HDG(T) - True Heading^^
TRK - Track
FPV - Flight Path Vector
and

Group2b
(also, Inertial/Rate sensor + computation)
ATT - Attitude (Pitch, Roll, Yaw)

We use the above data for 2 basic purposes:
(i) AIRPLANE FLIGHT & PERFORMANCE dependent on Group1 and Group2b data.
(ii) NAVIGATION dependent on Group2a data.

Some of what you state is possible, and already in use, using Inertial/Rate sensors coupled with computing modules, e.g. INS-like systems or AHRS (Attitude and Heading Reference Systems)

^^ The one prominent example of this, is the Heading (HDG). Inertial/Rate sensor systems have the property called gyrocompassing, ie their behaviour, as governed by the laws of physics, applied to the rotation of the Earth about it's axis, causes them to automatically align with True North as well as sense the Latitude at their geographic position. Thus the Magnetic HDG displayed on primary flight instruments in the cockpit is actually a synthetic Magnetic HDG provided by applying magnetic Variation from a database locally kept within the Nav Database or the AHRS itself. Such system architecture has done away with Magnetic compass/coil based master/slave systems etc.

Group1 parameters, represent the aerodynamics based flight of the airplane. Why use derived data rather than directly measured, proven and reliable systems, to fly a HDG, ALT and SPD? It is possible to derive a Group1 parameter such as the the AoA, by knowing the Relative Airflow, Pitch ATT from Group2a&b and the wing angle of Incidence. But this is (for want of a better word) a 'contrived' method of obtaining AoA when simple AoA sensor vanes are available and positioned by the the very airflow we seek to know about. The 2 main uses of AoA is to validate the pitot/static Speed data, and provide high AoA alerts. At the present 'state of the art' it is not practical, and it serves little purpose to substitute Air/AoA sensor Data with Inertial derived data except as an alternative in case of an ADC/AHRS system failure. Though we must note that with the advent of 'big data' and 'AI', ie Intelligent systems, this could change. Also, in recent history (last 2 decades), the failure of Group1 type system has proved fatal.

For the record, on some current airliners, when the Baro (Static sensors) measured Altitude is deemed unreliable, a GPS derived Altitude is available for display.

Answer 3

If I understood what an inertial navigation system (INS) is, it should be able to calculate speed and orientation of the aircraft.

In reading some articles about the 737 MAX I found out it is possible to derive AoA from other sensors but it will take more than the INS to get enough data to calculate this. AoA sensors tell the crew how the aircraft is moving through the air. Knowing where you are, the speed and direction relative to ground, and the position of the various control surfaces, is not enough to know AoA. It's going to take some measurements of the air movement as well. I don't know precisely what is the minimum required for this but an educated guess would be air pressure and air speed.

INS will tell you position and movement in relation to the ground, not in relation to the moving air.

If this is possible, why do airliners solely rely on sensors to measure speed and angle of attack (AoA)?

Part of the problem is that getting a new computer certified by the FAA is really hard. It's simply easier to get redundant sensors certified than the computers needed to calculate this information. This makes sense on some level because if we are to assume that hardware fails then how are we to know a calculated value is any better than a measured value?

As I recall Boeing has used AoA sensors on both sides of the aircraft, up near the flight deck, for some time now to provide redundancy. AirBus has used three for a long time, two near the front like Boeing and the third closer to the tail.

With Boeing there's an AoA disagree warning if the two sensors are outside of an allowed error margin between the two. While it is certainly helpful to know when the AoA reading can no longer be trusted it is more helpful to know what the AoA measurement should be. This is why AirBus has three sensors, a single sensor failure will not leave the crew without this information. With two sensors and a disagree warning the crew cannot know which reading to trust. With three sensors and a disagree warning then the two that agree can be trusted. If all three disagree then it's the same problem with two and one not working. This is not good but as I understand it not something to declare an emergency over, AoA measurement is not considered critical to safe flight.

The FAA, Boeing, and apparently a well trained crew, consider AoA information "nice to know" as opposed to critical to flight. This is why the FAA only requires two AoA sensors and an alarm that the reading is not reliable. Airbus, and some airlines, consider an AoA reading important enough that there are three AoA sensors on AirBus aircraft and this information is available to the crew. Boeing has AoA information available to the flight computer but it is not available to the crew unless the airline asks for that information to be displayed.

Because of the 737 MAX crashes it sounds like all future passenger aircraft will be required to have AoA available to the crew, and triple redundant sensors. AirBus already has three AoA sensors so they likely have little or nothing to do to comply with this requirement. It will be interesting to see how Boeing complies. I suspect all 737 MAX aircraft will have three AoA sensors before allowed to return to flight. Other Boeing aircraft could have another AoA sensor added or they can have a "virtual" AoA sensor by using information from other sensors and calculating the AoA value. At least one Boeing engineer thought it possible and beneficial to have a virtual AoA sensor when the investigation into the 737 MAX MCAS problem was going on.

What it comes down to is that the FAA, airlines, and flight crew all know that hardware can fail. Even if we assume software is flawless it depends on working hardware to function as it should. Whatever sensors that could be used to calculate AoA could fail so there would have to be some redundancy somewhere to detect this failure and still provide critical information to the crew. If we were to consider AoA critical to flight then having at least one AoA sensor would be wise to verify the calculated value. Having two AoA sensors means one AoA sensor can be lost and still have a verified calculated value. Having three AoA sensors means not needing to calculate the AoA value, it means having a way to verify each against the others.

Aircraft have had two of every flight critical sensor for a very long time. This left the crew with confidence in their measurements so long as everything agreed. Because some of this information overlaps it's generally possible to find which sensor is wrong if there is a single sensor failure, or no two sensors of the same kind fail. With a capable enough computer the process of eliminating a failed sensor can be automated, but the computer can fail too. This means two flight computers and three of some sensors.