3

The preliminary report of the ET302 crash included some data from the Flight Data Recorder. However, the position of the trim cutoff switches is apparently not among the recorded values.

Given that storage is so incredibly cheap nowadays compared to the price range of an aircraft, why don't manufacturers decide to record every single possible sensor value (including all switches in the cockpit) into Flight Data Recorders? Is this a purely economic decision to not do too much more than required by regulations? Or is there some other reason for it?

Florian
  • 3,839
  • 2
  • 19
  • 43

3 Answers3

2

I'm not an expert on FDRs, but I do have experience of hardware and software development in a regulated market (medical devices), so might be able to offer some insight. Here are a few considerations that occur to me:

  • Development cost: Adding storage to a FDR isn't like installing new DIMMs in your workstation. As well as the design work, the changed device must undergo new verification and validation cycles, the supply of the new components needs to be assured, and the manufacturer must continue to support existing products as well as a new one. The V&V may well include accelerated-life testing and other expensive activities; it's unlikely to be just a few days' work.

  • Customer demand: Aircraft operators require an FDR that satisfies regulatory requirements; it's a "tick-box item". Consequently, that's the requirement that aircraft manufacturers demand of their FDR suppliers. If there's no demand for extra storage (either for an extended loop time or for a greater number of sensor inputs), then there's no money to be made by offering that as a feature.

  • Interface limits: I don't know enough about FDRs specifically, but the input interfaces may well have a bandwidth limit which precludes providing more readings without providing further connectors (which then has a big impact on development costs; see above).

Toby Speight
  • 1,101
  • 11
  • 22
  • And yet many FDRs do record many more parameters than required by regulations - even the newer regulations only require a few hundred parameters, while many newer FDRs record thousands of parameters. – Vikki May 30 '19 at 02:02
  • Fair enough (I said I'm not a domain expert). I think my observations stand, but are probably not the whole story. – Toby Speight May 30 '19 at 06:58
0

Monitoring should not slow down systems updating data and data transmission between systems. Data storage is easy, but monitoring is expensive it might increase wiring, and overall load

user40476
  • 1,772
  • 9
  • 27
-1

Because storage is the easiest thing to add. What you think should be easy actually requires monitoring for every single switch and device, which would add a lot of weight and complexity. In aircraft accident investigations, switches don't usually move from their positions during and after a crash. That is why there is no value in adding that to the FDR dataset.

Juan Jimenez
  • 12,884
  • 2
  • 25
  • 67
  • 3
    Your 1st sentence is missing its main part. The 2nd sentence is wrong for any switches which are evaluated by digital hardware. 3rd sentence: in many crash investigations you cannot even find the switches any more. – bogl Apr 05 '19 at 14:14
  • 1
    @bogl And, in many cases, the "switch" is an option on a display screen menu. – David Richerby Apr 05 '19 at 15:21
  • 1
    You two need to look at a real 737 Max cockpit. You have no idea what you are talking about... I have worked in avionics for 40 years. You? httpss://www.straitstimes.com/world/united-states/us-pilots-complained-at-least-5-times-about-boeing-737-max-problems-records-show – Juan Jimenez Apr 06 '19 at 15:09
  • @JuanJimenez Which of my statements is wrong? – bogl Apr 09 '19 at 06:28
  • @bogl All of them. – Juan Jimenez Apr 10 '19 at 07:03
  • 3
    Are you serious? – bogl Apr 10 '19 at 07:23