Aircraft: Angle of Attack (A0A) Sensor design

As a member of the electronics design community, I am angry and disappointed with the recent deaths of 346 people due to the Boeing MAX 8 design. I always say, “Follow the money” and you will likely find the cause of most problems. Whether it is a rush to get a product to market before your competitors, or an effort to make more money on product options, the result is often not going to be good.

Regarding technology, why is it that we have developed reliable cell phones that can sense movement, compass direction, and much more? Why is it that we have developed an amazing MEMS technology that is rugged as well as improving so many innovative electronic systems (See Reference 5 and my comments below)? Can’t we do something to make an Angle of Attack sensor, which relies on micro-electro-mechanical movement, more rugged? It seems like the present designs have such a relatively high failure rate.

Of course, having two AOA’s, one on each side of the aircraft for redundancy and comparison, is a good solution. However, the second AOA was deemed an option by Boeing. That has changed since the accidents. See Reference 9.

CNN has found that Boeing relied on single sensor for 737 Max that had been flagged 216 times to FAA

What is the Angle of Attack?

The definition of Angle of Attack in an aircraft is the angle between the aircraft wing chord line and the incoming air passing by the aircraft in flight.

Why is the AOA important?

An aircraft will stall, regardless of speed, if it exceeds a high AOA known as the “critical angle of attack”

Figure 1

The Angle of Attack (Image courtesy of Reference 1)

The Angle of Attack (Image courtesy of Reference 1)

Safe Flight Instrument Corp. has a number of AOA solutions—see them here.

MEMS AOA sensors

It seems that MEMS5 devices in an AOA design have only been used in Wind Tunnels and on Unmanned Aerial Systems (UAS) to date. I can’t find one that is in production.

Calibration of the AOA6

Reference 6 has a really nice article about AOA calibration using Kalman filters. In this case, calibration requires having reference data to compare with actual output measurements of the sensor to be calibrated.

Determining flow angles (See Figure 2) requires high-accuracy reference measurements of AOA using measurements coming from an inertial navigation system and an air data sensor, and analytically reconstructing them using calibrated airspeed measurements and inertial data and the use of a Kalman Filter that was implemented in MATLAB/Simulink (Check out the Design of a high angle of attack pitch mode control in MATLAB Simulink)

Figure 2

Definition of Flow Angles (Image courtesy of Reference 6)

Definition of Flow Angles (Image courtesy of Reference 6)

Please share your comments and expertise with our audience on this subject.


1 Angle of Attack Sensor using Ball Mouse Encoder Wheel, Sumantra Bhattacharya, Akshay Mule, Subhasis Bhaumik, 2016 International Conference on Intelligent Control Power and Instrumentation (ICICPI), IEEE

2 Boeing AOA discussion

3 Satcom Guru

4 Not just the 737: Angle of Attack Sensors have had problems

5 Development of Alpha Sensor for Unmanned Aerial Systems Mi-Hyun Park, Sung-Su Kim, Chang-Kyung Ryoo, Keeyoung Choi, and Choonbae Park Department of Aerospace Engineering, Inha University, Incheon, Korea, SICE Annual Conference 2008 August 20-22, 2008, The University Electro-Communications, Japan

6 Wind identification via Kalman filter for Aircraft Flow Angles Calibration, F. Schettini, G. Di Rito, E. Denti and R. Galatolo, 2017 IEEE

7 Method of Estimating Angle-of-Attack and Sideslip Angle Based on Data Fusion, Hao Long, Shujie Song, 2009 Second International Conference on Intelligent Computation Technology and Automation, IEEE

8 Attitude Determination Using a MEMS-Based Flight Information Measurement Unit, Der-Ming Ma, Jaw-Kuen Shiau , I.-Chiang Wang and Yu-Heng Lin, Department of Aerospace Engineering, Tamkang University, Tamsui, New Taipei City 25137, Taiwan, Sensors 2012

9 Doomed Boeing Jets Lacked 2 Safety Features That Company Sold Only as Extras, NY Times, March 21, 2019

16 comments on “Aircraft: Angle of Attack (A0A) Sensor design

  1. BillWM
    May 15, 2019

    AOA Sensors have to have part of the sensor out in the elements — This means they are subject to bird strikes(the Ethiopian case), Paint, Masking Tape, DeIcer, Ice, Hail, runway debris, and other effects — I am pretty sure that RTCA DO-160 does not address bird strikes, debris, de-icer, and hail — This leaves these into the TSO for the sensor which is RTCA/FAA/CAA/ICAO  — The real key is having the complete and correct requirements for the AOA sensor design team.


  2. Steve Taranovich
    May 15, 2019

    @BillWM—You're right. DO-160 tests salt, fungus, fluids, sand, dust….not bird strikes or mechanical damage in docking and undocking the aircraft. I just think we need to be more innovative in our design approach to ruggedize the sensor by minimalizing mechanical aspects of the product where we can. From what I gather, it's not an electronic failure, but more of a physical/mechanical/movement error from bird 'gunk' or physical damage failure to the AOA from docking and gate departures.

  3. BillWM
    May 15, 2019

    Maybe what would be best is redundant dissimilar AOA sensing methods for each sensor  — Perhaps one the Vane and Shaft like there is now, and the other perhaps vector based pressure ports on a fixed metal protrusion or some other non-blade and shaft method.


  4. Steve Taranovich
    May 15, 2019

    @BillWM—that's an idea—we need to have more possibilities discussed like that and then test and choose from there

  5. emma8374
    May 17, 2019

    nice work

  6. BillWM
    May 19, 2019

    Did some checking and these sensors and the airspeed sensors in the last 20 years are linked to an Airbus A300 incident off of Brazil, an A320 series Aircraft In Europe and the two 737MAX losses.   Seems worth researching — The FAA/CAA/NTSB human factors analysis that these are recoverable faults must be missing something —

  7. Steve Taranovich
    May 19, 2019

    @BillWM—-Interesting—I will look into this when I get a chance

  8. BillWM
    May 21, 2019

    The one issue with the MEMS sensor that comes to mind is for a wing like an airliner wing with all the lift augmentation surfaces deployed in the high lift mode the required acuracy for the AOA sensor aproaches 0.1 Deg or less.   A smaller lighter plane/UAV appearantly does not require the accuracy (65kts takeoff speed vs 170kts takeoff speed)  — still seems worth researching but there is not much in the NASA / FAA budgets for aerospace R&D

  9. Steve Taranovich
    May 21, 2019

    @BillWM—Nowadays, due to budget constraints, NASA is working with private contractors like SpaceX, Blue Origin, and more. Why can't the FAA work with private contractors for something like this?

  10. BillWM
    May 21, 2019

    The real issue is in the air transport industry most of the prime and sub contractors have been “leaned out to the point of detonation” and the corporate budgets for R&D reduced — This is one of those things that would save airlines money in insurance costs over time, but is hard to come up with an FAA grant to a corp, or even university, where in the UK there is about 8 million pounds annually for companies and universites for things like this and similar in the EU.


  11. Steve Taranovich
    May 21, 2019

    @BillWM—-That L-1011 sensor looks to be much better protected against 'bird strikes' and possible taxi damage during arrival and departure at the gate

  12. BillWM
    May 21, 2019

    There are dozens of different AOA sensor designs used on different aircraft — my thinking is with two pairs of two different sensors one would not have all affected in an incident

  13. Steve Taranovich
    May 21, 2019

    @BillWM—you are correct–I say “Always follow the money!”—-Boeing made the second sensor optional and the airlines had to pay more for it! I believe that having two different sensors would be even more reliable, but I am not completely sure about that.

    Maybe we could get some Reliability people to chime in on that

  14. MWagner_MA
    May 22, 2019

    While not an expert in reliability engineering, I have had some training in it as you most likely have had.  My real fear is that the focus on software systems is overshadowing and minimizing the effort required to produce hardware that is reliable and up to the task of being mission critical capable.  How could an FMECA been allowed to pass muster with a single sensor!  Anyone with ANY training in FMECA/FMEA analysis would have stopped that project in its track.

  15. Steve Taranovich
    May 22, 2019

    @MWagner_MA—we appreciate your input and I would have to agree. It's just common sense when people's lives are at stake that a manufacturer should have redundancy for safety of these precious lives. And what about the FAA? Shouldn't they have pushed Boeing to have redundancy, as a safety standard and not an extra paid option, in a component that has shown past problems?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.