When All You Have is a Hammer…


Why We Automate Failure
A recent post on the interface issues surrounding the use of side-stick controllers in current generation passenger aircraft led me to think more generally about the the current pre-eminence of software driven visual displays and why we persist in their use even though there may be a mismatch between what they can provide and what the operator needs.

Ideally (or idealistically) in interface design one would start with the needs of the user and work forward to a specific implementation, be it tactile, auditory, visual or a mix. However it seems to me that in most cases we’re starting with the end in mind, e.g. a visual and symbolic display of information then shoe horning the implementation to fit the predetermined solution. My question is why?

I think the answer is simply that this is the technology that has evolved as the ‘front end’ for software. Going back to the earliest days of the computer industry there has been a preoccupation with the visual display of symbolic information and certainly this remains the preccupation of industry and academe today.

Now for many tasks visual displays are suitable, but a problem arises when this becomes a knee jerk single solution to all interface problems regardless of whether this is appropriate. We see this in the migration of information from what were traditionally sensomotoric (tactile) modes of input, such as the traditional stick shaker in aircraft cockpits, to visual and auditory channels requiring higher levels of cognitive function. Another classic example is the re-representation in ‘glass’ cockpits of analog displayed information as textual, thereby imposing greater cognitve workload upon the operator.

Ideally a robust systems engineering and human factors process should challenge the ‘automatic’ assumption that a visual display is the only way in which information can and should be provided to an operator. Unfortunately it seems that this is one sacred cow to big to sacrifice.

I guess this really is a case of ‘if all you have is the hammer, then everything starts to look like a nail’…

3 responses to When All You Have is a Hammer…


    An interesting thought you have there. My own take has always been that it fundamentally reflects the needs of the legal community, especially the dichotomy between the manufacturer and the operator. Manufactures have a desire to present information to the operator (the pilot as the operator’s representative) in a way that clearly shifts the legal onus away from the manufacturer. For the judge looking into legal liability for AF447 a stick shaker is an esoteric instrument but an aural or visual warning is much easier to comprehend. There is a desire to frame issues like information overload as primarily training issues and hence implicitly operator problems.

    People in the world of computer security often talk about “security theater” and I have been reflecting recently on the extent to which we can talk about “safety theater”. I would classify any change in a system that superficially appears to improve safety but actually doesn’t as safety theater. I think one example of safety theater is GPWS. The bias towards the visual display of information might also be another example of it.



    Why did three highly trained & experienced pilots crash a non-defective operating aircraft into the ocean? It took 13 long minutes to fall from altitude. More than enough time to recover from loss of control!!
    When you have non active shared side-stick controls with out stick shakers to take you back to your first stall you lose all your instinctive perspective to push the nose down, increase air speed, gain control & fly the damn plane.

    Why did the senior pilot not take over @ 10min before the crash ??? All that was needed was basic flying.

    Perhaps Airbus has created a dangerous and defective product.


      Matthew Squair 07/05/2012 at 9:43 am


      I think unfortunately the answer is a little more complex than just ‘a defective product’.

      Aircrew, like other highly trained professionals, do not necessarily make no mistakes what happens is more that the types of mistakes they make changes as they become more experienced. So one should not presume that an experienced crew will make no errors.

      A loss of air data or unreliable air data appears to be one of the most difficult situations for air crew to deal with and even with stick shakers the outcome is not certain. For example in North West Airlines flight 6231 (another pitot icing accident) the pilots misinterpreted the stall warning of the 727 stick shaker as the high speed buffet associated with overspeed.

      However I also think that there are aspects of both the man machine interface and the crew training regime (and certification regime) that were contributory factors to the accident. The trick is to remember that crew performance is not soley mediated by the machine and is also a reflection of cockpit socio-dynamics, organisational culture and crew training as well as the skills, strengths and weaknesses of individual crew members.