CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 12:54 am

In the past few years we have had the unusual opportunity to see two all-new and very unique aircraft types introduced to the industry - First the A380 and more recently the 787.
  • Both aircraft introduced after significant delays from their respective manufacturers.
  • Both aircraft brought with them significant changes from predecessor aircraft
  • Both aircraft received significant on-site intro support from the OEM and from suppliers
  • Both aircraft introductions have hit rough spots resulting in service interruptions and downtime

Given the interest here in every hiccup these introductions face (see active threads on LO, UA and AI introductions and challenges), I thought it would be worth opening a thread to discuss generally what makes for a successful (or troubled) introduction.

Watching a new airplane introduction from within an OEM is an incredible experience. Boeing views readiness for EIS as a 3-part problem to solve:
  • 1. The airplane must be ready
  • 2. The support products & services must be ready
  • 3. The airline must be ready

Preparations for introduction of the type begin years in advance. In the case of the 787, there was significant staff and effort focused on service readiness already in place when I joined the program in 2004. As EIS nears, and the airplane and support products are completed, the focus shifts to #3 from the list above – making sure the airline is ready. This effort must also start well in advance of the airplane delivering.

Here are a few of the key things an airline must accomplish ahead of taking the airplane to ensure they are ready for the introduction:
  • Develop a plan with the local regulator to get the type added to the AOC
  • Implement a training plan for pilots, engineers and mechanics
  • Order maintenance & flight training simulators as needed to support the training plan
  • Order GSE tooling – much of which will be long-lead items
  • Order spare parts – much of which will be long-lead items
  • Pre-position parts and qualified people at stations
  • Prepare maintenance facilities (hangars/docks) to support the new aircraft
  • Secure & provision gates to accommodate the new aircraft

Airlines get ready for a new airplane with varying degrees of success. Some are diligent at placing orders for long-lead tooling and parts in time to have them in hand when the airplane delivers; others are not. Some heavily provision outstations with parts; others provision little or nothing. Some have large amounts of support staff trained in advance of the introduction; some introduce the airplane and operate into stations with minimal personnel trained on the type. Some do extensive route-proving and hands-on training with the aircraft after delivery; some do none. This list could go on and on.

Regarding this last point of route-proving after delivery of the first aircraft, here is a summary of 787 introductions I posted earlier in the active LOT thread:

Airline - Time from first 787 delivery to first revenue flight

UA.. 41 Days
JL.. 35 Days
LA.. 31 Days
NH.. 30 Days
LO.. 30 Days
AI... 13 Days
ET.. 1 Day

What is remarkable about this list is that NH and JA each had large engineering staffs in Everett for years during 787 development. These airlines understood the airplane as well as Boeing by the time the airplane first delivered, yet still took a month or more to get to know the airplane before they put it into revenue service. UA, which is the largest and arguably the most technically capable of operators on this list, was not on-site in Everett during 787 development and took a bit longer with the airplane before putting passengers on it. AI and ET, with similar 787 exposure as UA in advance of delivery, but which have much smaller technical staff and less past experience with new types, put the airplane into service MUCH faster.

I don’t know that there is a “right” or a “wrong” way to go about an introduction of a new type, but there is no question that how you go about it has a direct influence on the smoothness of the introduction. Operators who spend less on parts and training in advance of the intro, and who skip or abbreviate any route-proving efforts, reduce their up-front costs for the intro and begin making money with their new airplane sooner. The downside is they are likely more exposed to disruptions and extended return-to-service times. Airlines who invest more up front are effectively insuring themselves somewhat from the financial and publicity impacts of a more painful introduction.

My experience is with Boeing products and most recently with the 787, but we can add a note here about the A380 introduction as well. Like the 787, introducing the A380 also brings many new and unique challenges. Airbus did a magnificent job supporting the first A380 introductions with large on-site staff and excellent levels of support. Still, much of the responsibility falls to the operator in order to make the introduction a success. Malaysia took their first A380 in on May 27, 2012 and did not put it into revenue service until 5 weeks later – this for an airplane type which had its EIS 5 years ago. It doesn’t always take this long (Thai took 2 weeks) but I believe this underscores both the challenge of introducing a new type, as well as the fact different airlines can take very different approaches to solving this very complex challenge.

[Edited 2012-12-22 17:50:43]
 
User avatar
Stitch
Posts: 26638
Joined: Wed Jul 06, 2005 4:26 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 2:05 am

Great idea, CM.   
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 3:17 am

Quoting Stitch (Reply 1):
Great idea, CM.

Thanks Stitch. Hopefully we can have a good dialogue about a relevant topic.

Also, I just realized I left QR out of my original post...

Airline - Time from first 787 delivery to first revenue flight

UA.. 41 Days
JL.. 35 Days
LA.. 31 Days
NH.. 30 Days
LO.. 30 Days
AI... 13 Days
QR.. 10 Days
ET.. 1 Day
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 4:52 am

What I would like to understand better is how does an OEM go about testing of the aircrafts hugely complex software system and how do you train the airline to handle it later. The reason is that today's indication of an airplane (or high end car for that matter which today has 30 computers or more on board talking to each other) get almost all indications that you have a problem through this incredibly complex software driven system. For a car you drive on angered by spurious warnings but for a frame you have to take everything more seriously. I know this was a lot of the problem with the A380 initially, ie there was no real problem but your software fault checking system said there was.

Part of operating a type successfully must be what to do with all those fault indications, when it is a real fault beneath the software and it's sensors and where it is only a tolerance or timing issue that would go away where you to reboot the subsystem or the whole machine.

I know the car industry put in an awful lot of testing on their much simpler systems (vendor testing, rig testing, mule testing and test car fleets of say 50-100 cars driven all over the globe for a year or so before introduction) yet the high end cars have been very bug ridden for the last generations, once again the dealerships are badly prepared to handle this all.

I therefore ask my selves how can the airframe industry handle something which is zig times more complex with 2-3 representative full cabin test aircraft during a year before EIS?
Non French in France
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 5:09 am

Quoting ferpe (Reply 3):
I therefore ask my selves how can the airframe industry handle something which is zig times more complex with 2-3 representative full cabin test aircraft during a year before EIS?

There's actually usually only one full cabin test aircraft (the other test aircraft don't have full cabins for other reasons). The way they try to work around it is to start early with very detailed requirements and interface definitions, then test in a great deal of depth.

Every component is bench tested. Then there are sub-system rigs (e.g. the whole cabin, the whole flight controls, the whole display system, the whole power system) where all the components of that system show they can play nice together, then you have the "iron bird" rigs of which there may be only one but it's typically active years prior to first flight, and finally the test aircraft.

A big part of the flight tests, in addition to accomplishing the test points of the day, is simply operating the aircraft. Squawks get written up on flight tests just like in normal service and they have to be troubleshot (without the benefit of a lot of prior experience or, usually, a fully developed set of manuals). This provides some self-selection in troubleshooting because things that are frequent recurring issues not only raise a big flag for something that needs fixing prior to EIS, they totally screw with test productivity so there's huge internal pressure to resolve those problems. Since OEM's don't have to operate under Part 121 and can operate as Experimentals, there are a lot more solutions available for these kinds of problems but they still have to be dealt with.

The final ultimate test is the Function & Reliability series...this is the last cert test of any major consequence prior to type certification. The details are negotiated with the regulators way in advance but the regulators can dictate any test point they want to see based on the cumulative experience with the aircraft up to, and during, the F&R testing. F&R testing basically consists of operating the airline as an airline would...you have to use the EIS procedures and manuals for operation and maintenance, you run proper dispatched flight plans, you have to do maintenance per the AMM, you have to do MMEL dispatches, etc. F&R flights have, at minimum, two regulator pilots (one from the Certification group that issues the type certificate, one from the Aircraft Evaluation Group that approves airline OpsSpecs) plus usually a regulator engineer, plus the OEM test crew. F&R is typically about 300 block hours per engine/airframe combination.

Tom.
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 7:53 am

Quoting ferpe (Reply 3):
Part of operating a type successfully must be what to do with all those fault indications, when it is a real fault beneath the software and it's sensors and where it is only a tolerance or timing issue that would go away where you to reboot the subsystem or the whole machine.

This is a real challenge with a highly integrated aircraft. It's not just a matter of getting the gains dialed in on when to set a message and when not to. Because modern aircraft systems are so intertwined one with another, a fault in one place can "cascade" through multiple systems, causing knock-on effects in multiple systems. When this happens, there may be dozens of fault indications appearing at once for multiple systems... all from a single failure.

To counter this, the airplane has onboard computing systems into which all system parametric data and monitored faults are fed. On the 787, we call this computing system the "Central Maintenance Computing Function" or CMCF. The CMCF employs complex modeling of how any given fault or combination of faults will effect all airplane systems in any given operating configuration. This modeling works with a combination of logic trees and algorithms. It is not unlike the 787's EICAS system in this regard, and in-fact the CMCF is working with much of the same data, just reporting at a much finer level of detail and with information which is targeted at maintenance & engineering rather than the flight crew.

The CMCF logic is first built based on the engineering knowledge of the airplane, but as systems are built and integrated in the lab, the modeling is constantly tweaked and expanded to included new learning about how the systems interact and effect one another. By the time the first airplane is built and powered up, the CMCF is functional and reporting faults. Lots of them. Too many of them. Much of what is reported initially are false positives or "nuisance messages".

As soon as the first airplane is built and powered, enormous effort goes into perfecting this logic to stop the nuisance messages. Nuisance messages will still occur to some extent even after the airplane enters service (one of the 787's first revenue flight cancellations was the result of a nuisance message which would not clear), but the modeling will likely continue to evolve for several years until all those kinks are worked out.
 
SCL767
Posts: 2812
Joined: Sun Feb 26, 2006 6:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 8:37 am

Quoting CM (Reply 2):
LA.. 31 Days

LAN took delivery of CC-BBA on 31/08/12. The first revenue flight was on 01/10/12 on the SCL-EZE route. LAN operated up to 12 weekly 787 flights on the SCL-EZE route during October. CC-BBB was delivered on 25/10/12 and on 01/11/12, LAN launched daily 787 service on the SCL-LIM route. CC-BBC was delivered to LAN on 07/12/12 and has been deployed on either the SCL-EZE and/or SCL-LIM routes. LAN decided to deploy the 787s on regional routes for the first three months since delivery in order for crew familiarization purposes and for ground crews to get used to servicing the 787s before they are deployed on long-haul routes from SCL and LIM. Boeing has been supporting LAN for over a year before LAN even received its first 787 in various ways and is still offering dedicated support to LAN. LAN is extremely happy and exuberant with the 787's performance. Next month, LAN will initiate long-haul flights with the 787s on the SCL-LAX and SCL-LIM-LAX routes. After CC-BBD arrives in late March, LAN will operate the 787s on long-haul routes from SCL and LIM to LAX and MAD/FRA.
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 9:44 am

@Tom and CM, thanks for informative answers.

Re fault indications and their root cause, in your experience, when a new model enters service how many of the reported faults are really faults after the replaced unit/subsystem has been analyzed in the labs.

Now that a unit does not show the reported fault can have several causes, either there was a false warning or the fault was caused by the unit interacting with it's environment and that specific environment will be missing once analyzed in the labs. Of the latter faults types, ie triggered by the specific plane environment rather then being a subsystem fault, what is the ratios there as well?
Non French in France
 
User avatar
KarelXWB
Moderator
Posts: 26968
Joined: Sun Jul 15, 2012 6:13 pm

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 9:45 am

Quoting CM (Thread starter):
My experience is with Boeing products and most recently with the 787, but we can add a note here about the A380 introduction as well. Like the 787, introducing the A380 also brings many new and unique challenges. Airbus did a magnificent job supporting the first A380 introductions with large on-site staff and excellent levels of support. Still, much of the responsibility falls to the operator in order to make the introduction a success. Malaysia took their first A380 in on May 27, 2012 and did not put it into revenue service until 5 weeks later – this for an airplane type which had its EIS 5 years ago. It doesn’t always take this long (Thai took 2 weeks) but I believe this underscores both the challenge of introducing a new type, as well as the fact different airlines can take very different approaches to solving this very complex challenge.

The first Malaysia A380 was a little bit odd, they took delivery of the frame on May 27 and flew it to KUL for crew training, then flew it back to TLS to apply the final livery and then flew it back to KUL again to prepare it for EIS.

Emirates put their new frames one day after delivery into service, but with almost 30 whalejets in the fleet they know how to do it.
What we leave behind is not as important as how we've lived.
 
hkcanadaexpat
Posts: 4019
Joined: Mon Dec 10, 2012 3:33 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 10:20 am

Quoting SCL767 (Reply 6):
After CC-BBD arrives in late March

No chance CC-BBD will arrive in March. It is scheduled to begin final assembly in Charleston in late March. CC-BBE will begin final assembly in Everett in late April. We might see CC-BBE be delivered before CC-BBD depending on pace pick-up at Charleston. Both should be ready for delivery in July/August 2013.

Cheers
A
 
SCL767
Posts: 2812
Joined: Sun Feb 26, 2006 6:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 11:04 am

Quoting hkcanadaexpat (Reply 9):
No chance CC-BBD will arrive in March. It is scheduled to begin final assembly in Charleston in late March. CC-BBE will begin final assembly in Everett in late April. We might see CC-BBE be delivered before CC-BBD depending on pace pick-up at Charleston. Both should be ready for delivery in July/August 2013.

This is why LAN has a contingency plan to temporarily deploy the A-340s on the SCL-LIM-LAX route from 02APR13 through 31JUL13. After CC-BBD and CC-BBE are delivered, two more A-340s will be phased out of LAN's fleet as LAN moves towards operating a widebody fleet consisting solely of Boeing a/c. LAN will also receive 5 new B-767-316ERs next year to ensure that the carrier can continue with its growth plans.
 
User avatar
autothrust
Posts: 1468
Joined: Sun Jun 11, 2006 8:54 pm

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 11:15 am

Quoting CM (Reply 5):
This is a real challenge with a highly integrated aircraft.

Exactly, almost all relability issues on the A380 were sensors or electronic.

Even for Lufthansa Technics there was a big learning curve with the A380. With the OIT and OMT you have access to any function of the aircraft and thousends of parameters. Further you can control
and monitor its integrated modular avionics system via a single interface (OMT)

Quoting CM (Reply 5):
The CMCF employs complex modeling of how any given fault or combination of faults will effect all airplane systems

On the A380 it's called Aircraft Condition Monitoring System (ACMS) AFAIK.
Flown on: DC-9, MD-80, Fokker 100, Bae 146 Avro, Boeing 737-300, 737-400, 747-200, 747-300,747-400, 787-9, Airbus A310, A319, A320, A321, A330-200,A330-300, A340-313, A380, Bombardier CSeries 100/300, CRJ700ER/CRJ900, Embraer 190.
 
swallow
Posts: 182
Joined: Wed Jul 11, 2007 12:23 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 12:45 pm

On the A380, both SIA and EK believed the warning system was too sensitive. TC is on record as saying the software was over complex, generated several false positive warnings, and reduced the dispatch reliability from 98.5% to 96-97%.

In such instances, does the OEM dial down the sensitivity of the ACMS with time after they have more in-service experience with the type?

Or do the crews learn which red lights to ignore?
The grass is greener where you water it
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 4:38 pm

Quoting SCL767 (Reply 6):
Boeing has been supporting LAN for over a year before LAN even received its first 787 in various ways and is still offering dedicated support to LAN.

I worked closely with LAN in preparation for their first 787. I can state unequivocally that outside of NH & LJ (who were embedded on the 787 program) LAN was by far the most prepared operator to take delivery of the 787. LAN's competence and dedication to being prepared for the airplane really blew me away. Hats off to them.

Quoting swallow (Reply 12):
Or do the crews learn which red lights to ignore?

In a Boeing, there are many different types of message. The type of message tells you whether you can dispatch or not. Obviously, ignoring any message is not a good idea, as they are being set for a reason:

In EICAS there are many types of message:

>> Warning - No go
>> Caution - No go
>> Status - Continue with the current dispatch (if engines are started) but must be addressed before the next flight.
>> There are also Advisory & Memo messages in EICAS

In the maintenance system:

>> Maintenance Messages - Tells you which LRU has a problem - Pilots don't look at them
>> Fault Codes - Tells you exactly what the airplane systems are reporting to the CMCF - Pilots don't look at them
 
2175301
Posts: 1614
Joined: Wed May 16, 2007 11:19 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 4:45 pm

Concerning fault detection. I am sure that airliners start with the same basic system in Power Plant Control systems where the "First Out" is detected and displayed with a different alarm (example - a whole series of alarm annunciators may light up for intertwined/cascading logic systems and problems; but, the "first out" annunciator for the problem that started the sequence is flashing on and off instead of just being lit).

20 years ago high speed printers kicked in and printed all the faults along with the time of initiation (to the millisecond); and I can remember laying out 20 ft long printouts that covered a plant trip that took 5 seconds to figure out what went wrong where. Nowadays computers sort all of that and display just the initiation events for the major system changes and that 20 page printout might only be 3 or 4 lines on a screen as different major systems shut down and the flashing annuciator buried in the middle of 30 other annunciators indicating an alarm condition.

Airplane cockpits do not have room for all of the annunciator panels that we have in power plants - but I am sure that the basis of the systems is the same.

Have a great day,
 
Sheridan125
Posts: 32
Joined: Sat Jun 30, 2012 5:54 pm

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 6:04 pm

Sorry CM. I do not like to contradict fellow a.netters but your post is just motherhood stuff. I was part of a team a few years ago which introduced 2 new aircraft types within 1 month. Both entered service on the required days. They were very different types. What was absolutely crucial was the selection of the management teams, particularly the engineering manager responsible for the despatch team, the fleet captain, the flight training manager and the Fleet Manager Technical (a pilot). Get these right and it will always work. Liaising with certification authorities, handling paperwork etc is what these people do well. Get these appointments wrong and the whole process becomes risky.
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Sun Dec 23, 2012 9:40 pm

Quoting Sheridan125 (Reply 15):
Sorry CM. I do not like to contradict fellow a.netters but

Not sure there's much contradiction. We're in agreement that getting the right people in key positions is critical to a successful intro. However, if an operator fails to order parts, train personnel, or otherwise invest in the intro, no management team will save you. You cannot manage your way around not having the right tool or missing parts which have months-long lead times. I've seen operators with great people suffer complete operational breakdowns because those great people were not put in place early enough and not permitted to plan and invest in the intro in the way they knew they should.
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 1:20 am

Quoting ferpe (Reply 7):
Re fault indications and their root cause, in your experience, when a new model enters service how many of the reported faults are really faults after the replaced unit/subsystem has been analyzed in the labs.

Most faults don't result in pulling the unit/subsystem to a lab...on the modern highly integrated stuff with BITE, you run the ground test and I'd say that the fault clears about 75% of the time with no further issue. It's only if you can't clear it on the ground that you pull the box. If you actually get that far, I'd say it's 90%+ that there's actually a fault in the component.

Quoting ferpe (Reply 7):
Of the latter faults types, ie triggered by the specific plane environment rather then being a subsystem fault, what is the ratios there as well?

I haven't run into that much...there are a couple of niche problems I can think of (a particular type of valve actuator is notorious for this) but, generally, if a component is kaput it stays that way. Intermittent defects, in my experience, are almost always wiring faults so if you pull the box and test it it will come back fine (because it is). I've spent many nights ripping EE racks apart trying to find that one @#%[email protected][email protected]%^ crimp that broke.

Quoting swallow (Reply 12):
In such instances, does the OEM dial down the sensitivity of the ACMS with time after they have more in-service experience with the type?

If that's actually the right fix, yes. In many cases it's a matter of adjusting counters or time thresholds.

Quoting swallow (Reply 12):
Or do the crews learn which red lights to ignore?

Flight crews should never ignore lights. Anything that rises to the level of crew action required should be taken seriously because any one flight crew indication can have tens or hundreds of potential causes underneath it and they're not all going to be nuisances. Mechanics get good, very fast, at knowing which warnings actually mean you need to do something and which are nuisances and just need to be reset so frequent offenders can usually be dealt with very quickly by maintenance. But they still need to be dealt with.

Tom.
 
User avatar
7BOEING7
Posts: 3039
Joined: Fri Oct 12, 2012 5:28 pm

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 2:22 am

Quoting tdscanuck (Reply 17):
I haven't run into that much...there are a couple of niche problems I can think of (a particular type of valve actuator is notorious for this) but, generally, if a component is kaput it stays that way. Intermittent defects, in my experience, are almost always wiring faults so if you pull the box and test it it will come back fine (because it is

In my experience on pre-delivery airplanes, from the late 70's to the early 90's if you had a squawk it was generally a bad box. Since then it's the wires. The "black boxes" are extremely reliable these days--some "infant mortality" but not much.
 
Unflug
Posts: 728
Joined: Wed Jan 11, 2012 5:25 pm

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 4:31 am

Quoting CM (Reply 16):
However, if an operator fails to order parts, train personnel, or otherwise invest in the intro, no management team will save you.
Quoting Sheridan125 (Reply 15):
What was absolutely crucial was the selection of the management teams, particularly the engineering manager responsible for the despatch team, the fleet captain, the flight training manager and the Fleet Manager Technical (a pilot).

To me it sounds you are pretty much in agreement. You have to order parts and train personnel in time, and the right management team will make sure that all is done in time.
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 9:02 am

Quoting tdscanuck (Reply 17):
on the modern highly integrated stuff with BITE, you run the ground test and I'd say that the fault clears about 75% of the time with no further issue. It's only if you can't clear it on the ground that you pull the box. If you actually get that far, I'd say it's 90%+ that there's actually a fault in the component.

That I get, the BITE (built in test) tests the unit in a well defined state and in a certain logical order. And that 75% of the cases the BITE goes through without showing a fault is what I expected, ie the fault indication was due to something not foreseen (as the BITE tries to go through all relevant parts of the unit and uncover all preconceived possible anomalies). The likelihood is now large that the fault indication is due to a not foreseen or tested combination of interacting events. How does one trace this down, are there signal state dumps for the last minutes of buses trafic etc that gets recorded for all fault indications (there should be) and that then routinely gets sent to the OEM together with the fault indication trace so he can get more real life data?
Non French in France
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 8:22 pm

Quoting Unflug (Reply 19):
To me it sounds you are pretty much in agreement. You have to order parts and train personnel in time, and the right management team will make sure that all is done in time.

We do agree. The point I was making (and perhaps was too cryptic about) is that some 787 operators have put great intro management teams in place but at the sr. executive level have failed to authorize the necessary expenditures to be ready for the airplane to arrive. When an intro team is handcuffed in this way, the intro will be very rough even with great people running the operational aspects of the intro.

Quoting ferpe (Reply 20):
That I get, the BITE (built in test) tests the unit in a well defined state and in a certain logical order. And that 75% of the cases the BITE goes through without showing a fault is what I expected, ie the fault indication was due to something not foreseen (as the BITE tries to go through all relevant parts of the unit and uncover all preconceived possible anomalies). The likelihood is now large that the fault indication is due to a not foreseen or tested combination of interacting events. How does one trace this down, are there signal state dumps for the last minutes of buses traffic etc that gets recorded for all fault indications (there should be) and that then routinely gets sent to the OEM together with the fault indication trace so he can get more real life data?

This is why your fault indication system must be more than just logic-based BITE. Algorithms which can work with parametric data are needed as well. The airplane makes available on the data bus more than 100,000 system parameters for continuous monitoring. During development and flight test, all of these are captured and used to refine the systems monitoring and fault indications which enter service. The "Airplane Condition Monitoring Function" (ACMF) has the ability to capture and report around 5,000 of these parameters per second. When the airplane enters service, ACMF has an OEM defined set of parameters which are captured and made available to Boeing and the airline. The operator can modify this set of parameters and choose to add or change what data the airplane sends them. In addition, certain events will trigger a "snapshot" of relevant data, which can be sent to ground real-time. An airline can also trigger a real-time/on-demand ACMF report containing a custom set of parameters, which is a terrific tool for starting to troubleshoot a problem while the airplane is still airborne.
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 9:57 pm

Quoting CM (Reply 21):
An airline can also trigger a real-time/on-demand ACMF report containing a custom set of parameters, which is a terrific tool for starting to troubleshoot a problem while the airplane is still airborne.

Quite advanced stuff, how long has such facilities been available and what kind of bandwidth has the data channel to base? I understand ACARS has very limited bandwidth.
Non French in France
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Mon Dec 24, 2012 11:20 pm

Quoting ferpe (Reply 22):
how long has such facilities been available

The capability was first available on the 747-400, then ironically dropped from the 777, Boeing believing it was seldom used. Operators informed Boeing otherwise and it is now offered on all 747s, 777s and the 787.

Quoting ferpe (Reply 22):
what kind of bandwidth has the data channel to base? I understand ACARS has very limited bandwidth.

ACARS is really only bandwidth limited for HF & VHF transmissions. Most widebody aircraft now have SATCOM, if not a Ku band connection, effectively eliminating the bandwidth constraint, at least as far as ACMF reports are concerned. ACMF reports are text-based and generally only a few kb in size. ARINC charges a king's ransom for even very small amounts of data on their HF network, which is why size is a concern at all.
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Tue Dec 25, 2012 5:24 am

Quoting CM (Reply 23):
it is now offered on all 747s, 777s and the 787

So I understand it is an optional feature, what base facilities is part of the package to evaluate all these parameters and what training is required for the technical personnel?
Non French in France
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Tue Dec 25, 2012 7:42 am

Quoting ferpe (Reply 24):
So I understand it is an optional feature, what base facilities is part of the package to evaluate all these parameters and what training is required for the technical personnel?

The reporting features are all standard equipment with the airplane. From ACMF, an airline gets relatively raw/unprocessed parametric data. Some sophisticated operators will take this data as-is and will work with it themselves on the ground, some having developed quite impressive in-house systems to make that data useful for troubleshooting and prognostic evaluation of the airplane.

Boeing also offers a for-fee service called Airplane Health Management (AHM), which processes this data real-time, turning into operationally useful/actionable information. Airbus has a similar service called "Airman".

To all my friends here on a.net - Merry Christmas & happy holidays. Wherever you are and however you celebrate this season, I wish you all a terrific end of 2012 and all the best in 2013.   

Cheers!

CM
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Tue Dec 25, 2012 11:38 pm

Quoting CM (Reply 25):
Boeing also offers a for-fee service called Airplane Health Management (AHM), which processes this data real-time, turning into operationally useful/actionable information.

So if eg UA is a subsriber to AHM they would get real time evaluations of operational issues from B, in consequence that could mean B experts could have been involved in the recommendation to divert the recent frame? (more to show the principle of information flow and how real-time things can be then to restart any UA diversion discussions).
Non French in France
 
CM
Topic Author
Posts: 623
Joined: Wed Feb 08, 2012 4:17 am

RE: 787 / A380 - Managing A Successful Type Intro

Wed Dec 26, 2012 6:55 am

Quoting ferpe (Reply 26):
So if eg UA is a subsriber to AHM they would get real time evaluations of operational issues from B, in consequence that could mean B experts could have been involved in the recommendation to divert the recent frame? (more to show the principle of information flow and how real-time things can be then to restart any UA diversion discussions).

No. For the flight environment, Boeing would not set up a service which would presume to provide enhanced decision-making, beyond what the flight crew could do on their own. By design, everything a crew needs to make the right decisions autonomously is provided to them in the flight deck. It would be a very unusual circumstance (e.g. external damage) where Boeing (or anyone outside of the flight deck) could know more about the airplane's fitness for continued safe flight and landing than the crew themselves.

AHM is primarily focused on the maintenance side, although it has some performance functions as well (e.g. tracking airframe drag and fuel burn degradation).
 
User avatar
EPA001
Posts: 3893
Joined: Tue Sep 12, 2006 8:13 pm

RE: 787 / A380 - Managing A Successful Type Intro

Wed Dec 26, 2012 3:37 pm

Quoting CM (Reply 25):
To all my friends here on a.net - Merry Christmas & happy holidays. Wherever you are and however you celebrate this season, I wish you all a terrific end of 2012 and all the best in 2013.   

Wish you and everybody else here the same. This is again a fantastic thread which is so interesting to read.  
Quoting tdscanuck (Reply 17):
I've spent many nights ripping EE racks apart trying to find that one @#%[email protected][email protected]%^ crimp that broke.

I like your selection of "words". 
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Wed Dec 26, 2012 6:39 pm

Quoting ferpe (Reply 20):
How does one trace this down, are there signal state dumps for the last minutes of buses trafic etc that gets recorded for all fault indications (there should be) and that then routinely gets sent to the OEM together with the fault indication trace so he can get more real life data?

In addition to all the montoring stuff that CM already described, most of the boxes (or hosted functions in the case of stuff that lives on the CCRs) have onboard non-volatile memory (NVM) that contains essentially everything the box ever thought about in recent history. So if you do remove and replace a box, the supplier can dive into its guts and identify exactly why it thought it had a problem. This will often involved parameters at a lower level than the airplane monitoring functions can see. If indicated, the supplier can adjust the BITE software to cover that situation in the future so these types of errors tend to attrition out of the fleet with time.

Quoting ferpe (Reply 22):
Quite advanced stuff, how long has such facilities been available and what kind of bandwidth has the data channel to base? I understand ACARS has very limited bandwidth.

The system will rarely dump a lot of data down the pipe without being asked...it will initially just report something short, like a fault code. Maintenance can then, if they choose, respond to that fault by requesting the aircraft send additional data. This way the airplane only uses the bandwidth that maintenance actually needs.

Tom.
 
User avatar
EPA001
Posts: 3893
Joined: Tue Sep 12, 2006 8:13 pm

RE: 787 / A380 - Managing A Successful Type Intro

Wed Dec 26, 2012 10:27 pm

Quoting tdscanuck (Reply 29):
The system will rarely dump a lot of data down the pipe without being asked...it will initially just report something short, like a fault code. Maintenance can then, if they choose, respond to that fault by requesting the aircraft send additional data. This way the airplane only uses the bandwidth that maintenance actually needs.

Are the systems Boeing uses comparable or identical to the systems Airbus is using for this? Are they maybe even cooperating in developing such systems? I am asking this since (if I remember correctly) just after the AF447 crash Air France was able to tell the latest condition of the airplane and that it had no obvious and reported defects?

Is there a trend to gathering and transmitting more and more data to the airliners maintenance centers or are the operators pretty much satisfied with the level of data they can receive already today? Wouldn't it be a good thing, but maybe expensive, to have such systems installed as standard on all airliners?
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Thu Dec 27, 2012 6:16 am

Quoting EPA001 (Reply 30):
Are the systems Boeing uses comparable or identical to the systems Airbus is using for this?

The functionality is comparable. I have no idea if the actual implementations are comparable.

Quoting EPA001 (Reply 30):
Are they maybe even cooperating in developing such systems?

Not that I'm aware of, other than they whole industry is converging on standards for maintenance stuff of all kinds and that's extending into fault reporting too.

Quoting EPA001 (Reply 30):
I am asking this since (if I remember correctly) just after the AF447 crash Air France was able to tell the latest condition of the airplane and that it had no obvious and reported defects?

My recollection is that AF had quite a thorough ECAM and maintenance message trace from the aircraft before it stopped transmitting...not enough to know what actually happened (flight crew action is on the FDR, not the maintenance system) but enough to know what some of the systems effects were.

Quoting EPA001 (Reply 30):
Is there a trend to gathering and transmitting more and more data to the airliners maintenance centers

Yes.

Quoting EPA001 (Reply 30):
are the operators pretty much satisfied with the level of data they can receive already today?

It's all about cost/benefit. Additional data needs to pay for itself but, so far, it has. So as long as the cost is reasonable and the implementation works, operators always like more data than less.

Quoting EPA001 (Reply 30):
Wouldn't it be a good thing, but maybe expensive, to have such systems installed as standard on all airliners?

Yes, although retrofitting them can be challenging. I'm pretty confident they're going to be standard on all new types
going forward.

Tom.
 
ferpe
Posts: 2667
Joined: Tue Nov 30, 2010 7:44 am

RE: 787 / A380 - Managing A Successful Type Intro

Thu Dec 27, 2012 8:05 am

One part of the preparation for a new type is to understand it's ground equipment requirements and prepare your complete network accordingly. I think we can see some of those issues in the UA test flights thread, the 787 has higher demand for electrical ground power and will therefore tax a ground equipment fleet that has lived a comfortable life as yet by not being asked to deliver to spec or close to spec. In essence a new type will also show what procedures needs to be changed when it comes to ground equipment being fit or not for purpose and how it has to be maintained and tested.

To what extent does one specify the below spec variability of e.g. electrical carts and other ground equipemtn and what the frame should stand in terms of "yes we know it is below spec but that is how the worlds ground carts are and we have to design for it" and how does one test for it before EIS?

The variability in makes and how they are maintained and handled must be large! Given that the 787-8 is designed as a long hop crusader it must go to the most remote places on earth and get to know the most exotic specimen of the breed in question  .
Non French in France
 
User avatar
EPA001
Posts: 3893
Joined: Tue Sep 12, 2006 8:13 pm

RE: 787 / A380 - Managing A Successful Type Intro

Thu Dec 27, 2012 3:17 pm

Quoting tdscanuck (Reply 31):
Yes, although retrofitting them can be challenging. I'm pretty confident they're going to be standard on all new types going forward.

That is good to read. Thanks for your extensive and as usual very informative answer.  
 
Lofty
Posts: 660
Joined: Sun Mar 16, 2008 5:23 pm

RE: 787 / A380 - Managing A Successful Type Intro

Thu Dec 27, 2012 4:35 pm

Best way is not to be one of the first to intrduce the type. Let others win the race to be the launch airline and then see the issue which A or B will correct before delivery of your aircraft.

This is why BA do not do Launch Customer for aircraft types these days.
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: 787 / A380 - Managing A Successful Type Intro

Thu Dec 27, 2012 5:02 pm

Quoting ferpe (Reply 32):
To what extent does one specify the below spec variability of e.g. electrical carts and other ground equipemtn and what the frame should stand in terms of "yes we know it is below spec but that is how the worlds ground carts are and we have to design for it"

Boeing has a good public document that gives an overview of what's needed:
http://www.boeing.com/commercial/airports/acaps/787electrical.pdf

The Maintnence Facility & Equipment Planning manual had exhaustive detail but that's only available to operators. It includes things like:
Nominal & allowable voltage range
Nominal & allowable frequency range
Distortion factors (total and per phase)
Maximum continuous and peak power consumption
Thresholds for the bus protection system over/undervoltage and over/underfrequency trips

Quoting ferpe (Reply 32):
and how does one test for it before EIS?

You get a variety of GPU's and try them. This sort of happens by default since there's a lot of testing away from home base and you take what you can get when you're away from home. The power system guys can test the exact thresholds by using controllable power supplies but, as far as I know, that's usually done at a component level rather than a whole airplane level.

Quoting ferpe (Reply 32):
The variability in makes and how they are maintained and handled must be large!

It's appalling. Something very similar happened with the 777 introduction, since it was (for its time) also a giant ground power hog, and is actually less tolerant of low quality ground power. If you ask the FBO guys in a quiet moment which GPU's are good and which aren't they always know...some are dogs and some aren't. Some are fine but can't make rated power anymore...they're great if you don't load them fully but they'll die in a heart beat if asked to make rated power.

Tom.

Popular Searches On Airliners.net

Top Photos of Last:   24 Hours  •  48 Hours  •  7 Days  •  30 Days  •  180 Days  •  365 Days  •  All Time

Military Aircraft Every type from fighters to helicopters from air forces around the globe

Classic Airliners Props and jets from the good old days

Flight Decks Views from inside the cockpit

Aircraft Cabins Passenger cabin shots showing seat arrangements as well as cargo aircraft interior

Cargo Aircraft Pictures of great freighter aircraft

Government Aircraft Aircraft flying government officials

Helicopters Our large helicopter section. Both military and civil versions

Blimps / Airships Everything from the Goodyear blimp to the Zeppelin

Night Photos Beautiful shots taken while the sun is below the horizon

Accidents Accident, incident and crash related photos

Air to Air Photos taken by airborne photographers of airborne aircraft

Special Paint Schemes Aircraft painted in beautiful and original liveries

Airport Overviews Airport overviews from the air or ground

Tails and Winglets Tail and Winglet closeups with beautiful airline logos