MOQA / FDM maintenance predictive analytics

[peekay4]

Hi all, I'm trying to understand the use of FDM for maintenance operations (e.g., via Safran AGS) -- especially related to analytics to predict technical issues.

I haven't found much literature on this subject on the web. E.g., what parameters are captured in the FDM, what is the process to download & analyze the data, what types of technical issues can be detected or anticipated based on the data, etc.

(A related question is whether AOG desks use / can use aggregate analytics to make inventory decisions).

Any pointers & help greatly appreciated!

[YOUNGBUCK]

Hi all, I'm trying to understand the use of FDM for maintenance operations (e.g., via Safran AGS) -- especially related to analytics to predict technical issues.

I haven't found much literature on this subject on the web. E.g., what parameters are captured in the FDM, what is the process to download & analyze the data, what types of technical issues can be detected or anticipated based on the data, etc.

(A related question is whether AOG desks use / can use aggregate analytics to make inventory decisions).

Any pointers & help greatly appreciated!

Hi,

The data recorded varies from aircraft type but there are mandatory parameters which are recorded as a minimum, I believe these will be stipulated in cs25 but someone will correct me if not.
Data can be downloaded as a scheduled task or at request using a piece of kit. Data can also be sent via recorder wirelessly or via a memory card.
The data can be used by Engineering to detect excessive accelerations or speeds for the flight configurations etc.
I would say for AOG desk it would be better to use system data transmitted by acars type systems for spares positioning.

[peekay4]

Thanks Youngbuck.

I looked through the mandatory parameters however (as one would expect) virtually all of them seem to be parameters suitable for accident reconstruction -- things like speed, altitude, control yolk position, ice detection -- but not necessarily useful for maintenance actions?

I imagine there's much more engineering-type data available from the QAR? I can't find a comprehensive list of QAR parameters for (say) an A320 or B737.

Question re: ACARS, for something like engine monitoring. Does the system regularly transmit engine health parameters (even when there's nothing wrong) -- or will ACARS messages get generated only when anomalies are detected / certain thresholds exceeded?

Thanks for your help.

[woptb]

RR use live engine health monitoring (EHM) via ACARS/satellite and onwards via ground stations. Try RR website for more info. Would guess GE & others will offer the same service.

[YOUNGBUCK]

Thanks Youngbuck.

I looked through the mandatory parameters however (as one would expect) virtually all of them seem to be parameters suitable for accident reconstruction -- things like speed, altitude, control yolk position, ice detection -- but not necessarily useful for maintenance actions?

I imagine there's much more engineering-type data available from the QAR? I can't find a comprehensive list of QAR parameters for (say) an A320 or B737.

Question re: ACARS, for something like engine monitoring. Does the system regularly transmit engine health parameters (even when there's nothing wrong) -- or will ACARS messages get generated only when anomalies are detected / certain thresholds exceeded?

Thanks for your help.

For maintenance actions, interpretation is still required. For example on an airbus (if configured) : the aircraft has a ecam caution in flight, this will transmit via acars to a Maintrol dept or similar and from there they can arrange potential spares with AOG desk (simplified scenario)
With regards to engine parameters these are regularly transmitted and require analysis by a power plant dept or engine provider on a program such as aerobytes.
Often we get a request from the people who trend monitor to perform a maintenance action because they notice a negative shift in parameters.
EGT shift for example can lead to a compressor wash being requested.

[peekay4]

Thanks woptb and Youngbuck.

Do you know if the various monitored parameters are also used to make fleet-wide predictions (vs. trend monitoring of individual aircraft)?

E.g., perhaps fleet-wide (in aggregate) part X seem to be wearing down (or breaking down) faster than expected, and since part X has a long lead time, a purchasing decision is made sooner.

Have you seen anything like that?

[YOUNGBUCK]

Thanks woptb and Youngbuck.

Do you know if the various monitored parameters are also used to make fleet-wide predictions (vs. trend monitoring of individual aircraft)?

E.g., perhaps fleet-wide (in aggregate) part X seem to be wearing down (or breaking down) faster than expected, and since part X has a long lead time, a purchasing decision is made sooner.

Have you seen anything like that?

The example 100% happens but I'm not sure about the role of FDR systems in this instance. I do know evaluation of failed components is used to source components with long lead times in anticipation or part changes earlier than necessary as preventative maintenance.
This is a role of a reliability department / TC holder and is usually for airlines with cash!