SYMPOSIA

The Aviation Workshop was held on the afternoon of June 5th, 2003 as the final activity of the SAE Vehicle Recorder TOPTEC Symposium. There were two presentations followed by a discussion period which was open to any aviation recorder-related issues/ideas. There were approximately 33 people who attended the session representing all aspects of aviation including general aviation, air carriers, aircraft/recorder/component manufacturers, military, government, investigative, and pilot union's. The session commenced at 1330.

The first presentation, "Installation of Combined Voice and Flight Data Recorder", was given by Robert Swanson of FedEx Express. Mr. Swanson explained why FedEx Express chose to replace the traditional FDR/CVR configuration on their aircraft with a redundant CVFDR/FDR configuration. He also detailed some of the benefits resulting from the installation and shared the lessons learned from the experience. The presentation can be found in the proceedings.

The following were some questions from the attendees that arose from the presentation.

What was the start logic used for the FDR portion of the CVFDR?

• In FedEx's case, no start logic was used in the CVFDR for starting the FDR portion of the recorder
• This allows the capture of additional data that occurs at times when the stand-alone FDR is not recording
• Start logic varies from system to system
• Having no start logic for the FDR means that if there is power on the recorder, it is recording.
  • Not all parameters are valid before complete aircraft system initialization
  • Some parameters that may be significant to startup events are valid

What MEL relief was received as a result of installing the CVFDR?

• MEL relief for the FDR was extended by an FAA policy letter (PL-87, rev. 6) which no longer requires an aircraft to stop at the next maintenance station to fix the FDR which allows more flexibility to the operator.

Dennis Grossi, National Resource Specialist Flight Recorders, NTSB, gave the second presentation, "Quality of Flight Recorder Data - Filtered/Under Sampled Data and Post Processing Errors/Limitations". The talk focused on some of the effects of poor data quality and regulatory non-compliance on data collection and how accident investigations are hindered. The importance of valid post-processing of data was also stressed. Mr. Grossi's presentation can be found in the proceedings.

Some topics that arose from the discussion:

Ground Readout Equipment Certification

• Data being processed is being used in operational/maintenance decision-making. How do we ensure validity of the data since there is no requirement to validate readout software?
• Do we need it to ensure valid software/conversions/sampling?
• Is it feasible to require validation of a software product?
• Will regulatory compliance squash the flexibility of airlines using their data?

QAR Data stream needs to be in protected memory

• QAR data stream is usually better than required data stream
• QAR exists because of difficulty changing the mandatory stream
  • QAR stream does not have to be certified
  • Tools are available for owners to modify their own QAR data stream
• Flexible mandatory data stream could be used for both
  • Although FAA stated in RTCA FFDCC that the method exists for changing mandatory data frame without having to re-certify, manufacturers and operators are not convinced
  • Most FDAU manufacturers do not have tools for end users to re-configure their own box

Building data-map into recorder

• Could some of the documentation problems be solved by building the dataframe into the recorder?
• Version control - Edits could be uploaded and version number adjusted on powerup?

Data under sampling, latency, and filtering

• Under sampling occurs mainly when data for the FDR are acquired from systems that use lower sampling rates for a different purpose. (flight controls, cockpit display, etc)
• Under Sampling has its roots in several different areas
  • Designers are not aware that their system is going to be used to feed the recorder and are not aware that they need to have higher output sampling rates to meet the requirements
  • Integrators are not doing proper integration to ensure that the recorders are getting the correct number of fresh samples
  • Regulators are not checking to ensure that stale data values are not recorded
• Designers use different CFRs than operators and the FDR requirements are in the operations section of the CFR
  • Change CFR Part 25 to reference or contain the same recorder requirements in Part 121
• Once sample rates are reduced on flight critical systems, they will probably never get upgraded to FDR rate requirements due to the cost of re-certifying the flight critical system.
  • Higher sample rates could easily be integrated during initial design but it will never happen once a system is fielded
  • Sensor characteristics are based on critical system requirements and care needs to be taken to ensure that higher sample rate recordings are getting valid data
• It may be time to re-address sample rates for all parameters
  • Original sample rates were based on limited capacity of recorders
  • New sample rates should be based on the bandwidth of the parameter to be sampled w/o restriction due to capacity

AEEC FDR Working Group

• Working group for developing ARINC standard for FDR documentation
• Starting in August 2003 - Robert Swanson Chairman

RTCA Future Flight Data Collection Committee results
The RTCA FFDCC committee was generated by Chairman Hall and Administrator Garvey out of the 1999 recorder symposium. The committee met and created a report with recommendations.

• What, if anything, is being done concerning the results/recommendations of the RTCA report?
• Is it necessary that another group be formed to address those concerns?

The TOPTEC Aviation Workshop provided an opportunity for individuals to come together and discuss some of the current issues in the field of aviation recorders. This paper has attempted to document some of the ideas and concerns discussed during that meeting.