Tuesday, November 5, 2013

787 Glitch Statistics By Press AI Leads the Pack

It dawned on me awhile back that Air India has an inordinate



amount of glitching. You see them as false lights blinking, Panels dropping off on final approach, and faulty parts. So I looked at where AI Inventory came from in the 787 line of craft

Below are AI first 10 units and where they are built.

Everett                Charleston

# 1                            #3
#2                             #4
#8                             #5
#10                           #6
                                 #7
                                 #9

Totals  4                   6

Step one is to isolate a problem containing two factors. Customer proficiency and source of build. AI has reach a statistical number most people can relate to and come up with a broad assumption without tying any one glitch to a source. This is dangerous ground in sampling.


  • 98 aircraft built by  Boeing and delivered
  • 10 in sample size
  • 10% Sample Size From Boeing is The AI Fleet
  • 60% of all AI 787's are built in Charleston
  • 50% of Charleston's builds have delivered to Air India
  • 40% of all AI are built in Everett



More detail in stats:


  • 98- 787 are delivered representing 100% of the population tested.
  • 4- 787's from Everett = 4.1 % of Boeing population and 40% of AI 787 fleet
  • 6- 787's  from Charleston = 6.1 % of Boeing population and a 60%  0f AI Fleet


Now for some more number crunching without any empirical or actual glitch data.

The probability of glitches in Air India fleet should have 60% assigned to Charleston built aircraft and 40% assigned to Everett built aircraft. What becomes curious is that Charleston does not build 60% of Boeing's aircraft but some number ssouth of that, causing a weighted skewing in this example.

Charleston has delivered 12 out of the 98 787's in customers hands of which AI now owns 6 or half of the Charleston production. AI has the highest glitch rate publicized for awhile now. The main thing about statistic is to develop a trend or a predictive picture of outcomes in an manufacturing process.

Having had a lifetime effort wrapping my mind around statistics during performance auditing or error statistics you could develop conclusions based on the tests and raw outcomes. AI is a good case for this type of work. Several things need to happen at both ends of the test.. One is a Boeing Everett to Boeing Charleston comparison of delivered  Aircraft for its completeness and fault expectation, the second is Air India's ability to manage this new technology.

If a performance audit would determine AI is doing everything required operating its aircraft, then the shoe drops on the production source. If a majority of the glitches occur on the Charleston built aircraft where AI owns 60% of its fleet from Charleston, then Charleston has some serious maturing before its ready for prime time.. Boeing owns that data in the far reaches of its relational data base and is not public information. Completeness tests in auditing should be conducted on the process of building the 787. Something appears to be out of skew.

Summary:
  • 12.25% of all 787''s comes from Charleston
  • 60% of Air India is From Charleston, representing half of the Charleston 787 ever built and delivered.
  • Air India is a Leader in frequent Glitching


Now here comes the "Nail Biter":

Both Norwegian Aircraft come from Everett and have had a recent rejection inspection on its aircraft where Boeing came and fixed it, Anomaly?