787 Delays. Boeing or subs, whats your opinion? 3

[DHambley]

This thread is meant as a forum to discuss the issues contributing to the late schedule for the 787. Do you work for a subcontractor or for Boeing? What is your product? What are some of the items causing late delivery. What can Boeing do differently?

(I had mistakenly put this thread in a different subgroup. This is the correct Forum)

 

[tbuelna]

Boeing willfully put themselves in a position where they are dependent upon their suppliers. Both parties are to blame.

 

[IRstuff]

It's always the prime's responsibility to manage the schedule. Boeing has gotten particularly bad at managing extremely large activities, but that might be a symptom of projects being too large for anyone to manage adequately. Additionally, the problems appear to a combination of the mundane, lack of fasteners, and possibly more systemic, with the major subs tasked with jobs larger than they have experience and infrastructure for, since Boeing subcontracted major subsections of the plane to their suppliers.

When such large subsystems are subbed out, there's no room for second sourcing, and there's less visibility into any possible problems. Had the plane's 2nd tier assemblies been assigned to the main Boeing production line, many of the current problems would have been apparent sooner, and possibly solved sooner.

Unfortunately, such is the nature of large programs today. The level of complexity is substantially higher, which means that the prime's program manager and project engineers are essentially an extra layer away from potential problems and solutions.

It's not really clear that there is any solution. It used to be that a single chief engineer could understand and encompass the totality of an aircraft development and production. Today, each major system of an aircraft, or military vehicle, is so complex that chief engineers at the 2nd or even 3rd tiers have trouble coordinating the engineering activities, just because there's so many of them.

Just consider the most advanced landing gear. They used to be simply a bunch of mechanical components that were assembled together. Now, they have electronics, sensors, processors, actuators, etc. Some landing gear will even have motors to spin up the wheels prior to touchdown, to lessen the friction induced tire ruptures. These motors then act as brakes when the plane slows down. All this requires engineers versed in mechanics, physics, materials, electronics, control systems, sensors, etc., not just mechanics and materials.

TTFN

FAQ731-376

 

[prost]

Great post, IRStuff!

A new aircraft is an exceedingly complex thing to build. It is rare for an entirely new aircraft to be built on time and on budget (what's the old saw? "I can give you only two out of three: fast, cheap, or good."); it seems almost invariably, schedules slip, hundreds of millions more are spent than anticipated, customers get mad because of delayed deliveries.

Boeing doesn't have a monopoly on what looks like poor management of complex operations; Airbus's A-280 was years behind schedule, EADS A400 is also years behind schedule. In the case of the 787, you add to the usual headaches of a new aircraft the complications of dealing with more Tier 1 suppliers located around the world, and what you got was almost inevitable; schedule slips, cost overruns, and mad customers. I am sure Boeing is banking on the coordination problems being taken care of next new aircraft they build, and the result will be for them a leaner, meaner organization, more adaptable to fighting Airbus and the new threat coming from the PRC, which is starting to build a regional jet.

 

[GregLocock]

Does anybody think that the elimination of real protoypes might be something to do with their problesm?

I must admit I do somewhat buy into the idea that the problem with prototypes is that (a) you have to design it once for the proto and again for production, and (b) every problem you raise against a prototype gets dismissed by management as it is a non representative design.

On the other hand prototypes do allow you to measure loads in new systems and catch snafus with system integration, in particular.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[IRstuff]

No, not really. While there's something comforting in having a prototype, when it fails in testing, it's a LOOOOONG iteration to get it back in the air again. An FEA is back in the air in a matter of days or weeks. Additionally, what the FEA models do is to ensure that the actual qual units are much closer to qualifiable than they would be otherwise. A couple of MD-80 qual units wound up cracking their tail sections because of, oops, the planes were too long for a particular flare maneuver, something that could have been caught in FEA analysis.

Boeing appeared to have relatively minor issues with the actual qual units. It's the production ramp up that has been hiccupping, and had prototypes been built, they wouldn't have numbered enough to really stress the subs.

TTFN

FAQ731-376

 

[rb1957]

i agree with greg, that part of the problem is the rush to production. Structural issues (in my experience) tend to be easier to fix than systems problems. And systems seem to take longer to mature.

also, partners are a terrible "innovation". understandably, they have their self interest at heart; i've lived through several programs where there was a "bun-fight" between the design initiator and the partners (heck, i've thrown quite a few buns myself) that wouldn't've happened if the partners were a group within the design initiator.

 

[swall]

Wasn't the 777 done without prototypes? And didn't that program go fairly well?

 

[GregLocock]

I think BillBirch's first post in thread731-220877 is very relevant. If you take exceptional care and plan for it and include extra resources to flood problems, then you can accelerate your development/build schedule.

BUT, when you've done it once then the program managers assume that the extra care was unneccessary, and you can now meet the accelerated schedule with no protos with just the usual resources.

Not that I'm living through that nightmare at the moment. Oh no.



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[bf109g]

A totally irrelevant comment - whatever the reasons for the delays, it's holding up 767 cargo conversion feeder stock.

Which means that I'm going through a hurry up and wait at the moment.

 

[rb1957]

which one ? swall's or greg's ?

swall has a point that the 777 seems to have been a well run program. maybe they too a lot a carefull conservative approach, maybe they had better (more technical) management, certainly they had more involvement (less, smaller partner roles). and yes, they did it without prototypes ... you don't Have to use prototypes but they are good at risk mitigation; but maybe they demonstrated the new technology on rigs ahead of the main design.

i think greg has a point too, that a good outcomes makes you complacent.

the fact is that pert near every project is late and over-budget, but things seem to work themselves out in the end !?

 

[bf109g]

rb -

Everybody here is making good points, yourself included.

I was merely being self centered in my comment that the 787 delays are affecting my current project load.

 

[IRstuff]

The 777 had a fundmentally different production plan, which was more in line with the traditional practice of Boeing doing the bulk of the major assembly.

The 787 is structured around subcontracts doing most of the assembly, with only final assembly done at Boeing. There's probably a factor of 10 times decrease in final assembly scope at Boeing, and a likewise 10 times increase in scope at the subcontract.

TTFN

FAQ731-376

 

[rb1957]

and the subbies had a much bigger role in the engineering side of things

 

[DHambley]

I've worked for 3 subs who are building electrical systems. The paperwork demand from Boeing is incredibly higher than for any other aircraft I've worked on.
Some paperwork examples:
More analysis to support reliability than ever before.
New specification formats to comply with DOORS.

Also much more built-in-test (BIT) circuitry is designed into these systems than ever before. The failure-rate calculations look better on paper with all of this added circuitry. (Is that really better though? That's a huge topic alone).

 

[IRstuff]

It seems that one possible mode of design for failure is the subcontracting mechanism that's currently in vogue at Boeing. Traditionally, every piece of purchased hardware incurs a "wrap," that accounts for procurement, incoming inspection, stocking, economics, etc.

What Boeing has done on programs I'm involved in to avoid incurring wraps is to directly procure hardware but have the integrator do the integration of systems that he didn't procure or specify. What we then get, when we try to arrive at a mutually agreeable interface is, "Oh, that's not in OUR spec, you'll need to have Boeing change our spec."

There's probably a lot of that going on in the 787, based on the hgldr's comment.

TTFN

FAQ731-376