FDC
Where the Data Did Go…?
Dave Jesse on October 19, 2016
2

In my last two blogs I explained that we had identified a problem where small amounts of data were occasionally being lost with uQAR-equipped aircraft. After ten years of satisfactory operation, this was very puzzling. I also outlined how we built a test set using a Raspberry Pi computer and showed some sample test data.

I did promise an answer, and after about 30 tests of all the combinations of data cards and recorder modification states, totalling over 250 hours of testing, the answer is…

The Answer

It’s the data cards, but it’s not that obvious. We purchase data cards through a local supplier who specialises in solid state media. Over the last twelve years we have placed 114 orders with this company, and their knowledge of the products is excellent.

The card manufacturer they recommend for small capacity industrial grade Compact Flash cards changed the firmware in the card in two stages. The original and second variations were perfectly fine and never dropped a bit of data. The third variation, released in late 2015, was identified as the version which caused the data loss.

Now, the card manufacturer is not a small company, with a turnover of about half a billion pounds and a commitment to quality, holding ISO 9001, ISO 14001, QC080000 and ISO/TS 16949 approvals, and “all products meet the company’s rigorous quality standards.” Sadly, they just don’t work.

David and Goliath

Here is where the theory of supplier management can fall down. Flight Data Services’ total orders amount to less than one millionth of the turnover of the card manufacturer, and trying to get their attention is impossible.

The Solution

We have scoured the suppliers in Europe and America and found stocks of the older, but working, cards. This means that we can replace any faulty cards.

Better than that, as we know the cause of the data error we have built a test into our data transfer software that will identify any faults in service and advise the user to return the suspect data card for replacement. In this way, any faulty cards can be trapped. This is not ideal, as some data will have been lost in order to identify the problem, but at least we’re on the right track.

There is nothing like clarity in a warning message, and I think this is certainly clear enough:

warning

The Long Term

These recorders are excellent devices, but we are now aware of the limitations of the media in use. With a test set able to identify errors in the recordings we are well placed to check out the next generation of data cards before supplying them to our airline customers.

  • Al Gom

    Dear Dave,
    Congratulations for all the great job that you and your team have done with data from FDR system and for sharing all this knowledge and experiences with all the professional and enthusiastic aviation community in this forum…..
    One other hand, I’m wondering now if you are considering to contact L3 and let them know your findings about the data card issue. Maybe they could be interested in developing a modification or software update in their product to deal with this problem …
    This is a potential situation for the manufacturer since the original and 2nd variation datacards will be progressive and inevitably phased out to be replaced with the third variation and the commercial availability for the old ones will be less over the time. In any moment operators will be forced by the market itself to use the new and defectives ones…

  • Dave Jesse

    Dear Al Gom,
    Thanks for your kind words. We have shared this, both with L3
    and with our friends at Aerobytes as we know they supply the same type of
    recorder and their customers may have experienced the same problem.
    As you mention, the availability of media is, and has always been, a problem in the long term. We are seeing the same with Magneto-Optical disks which are almost extinct, but some operators are still using disks that would have been considered unserviceable back in the days when replacements were available.
    Dave