I was reading thie comment viewtopic.php?f=3&t=1437867&start=950#p21932263

Specifically this part -

[quote=patrickjp93]As an engineer myself, I think grandfathering and iterative improvement are perfectly valid, and it does NOT always make sense to start from a clean sheet[/quote]

I am failing to understand the bit shared above. I didn't want to pollute the thread as it has been too far back hence asking here. I use Debian and have also contributed a tiny bit. What I have seen in iterative methodologies that as changes happen, new tests are written all the time to see that things work coherently. An example is autopkgtest which is used and being used more and more by new and updated debian packages see https://wiki.debian.org/ContinuousInteg ... utopkgtest

What is interesting is that there is lot of interplay of hardware and software here as well. For e.g. I live on testing on sid and there are many others having different software-hardware architectures who update their systems as and when. Some might even jump from stable to testing as and when they like . So there is lot of testing and Q&A going on by the volunteers themselves. I know it's different but I was wondering if any of such practises could be used there or not ? Another question is should there be some hard limit to grandfathering or not especially in civil aviation ?

I have been hearing about grandfathering for quite sometime with the Max . Are there any other aircraft which have been notable or known for grandfathering their designs ?

https://en.wikipedia.org/wiki/Grandfather_clause

This isn't a software thing, it is a type rating that continues to be the same so airlines dont have to do months of training per pilot for the newer model. A320 NEO, A330 NEO, E2 jet and 777X will also fall into this category.

I guess a discussion could be around the regulations on what or how many changes can be made to retain the similar type rating.

Another factor is the ability, via grandfathering, to be subject to older (presumably less stringent) requirements in certification.

pune wrote:

I am failing to understand the bit shared above.

There is an old saying, better the devil you know.

The 737 built today is very different from the first 737. Whenever you build something from scratch it has unknown problems. Over time these problems should be resolved through in service experience, service difficulty report, or accidents.

Building on the baseline of that improved 737 means many problems have already been addressed, so you are starting from higher level of safety.

We see this in all sorts of engineering worldwide, you learn from your previous designs.

I must say never did I think Linux and this forum would ever cross over

Phosphorus wrote:

Another factor is the ability, via grandfathering, to be subject to older (presumably less stringent) requirements in certification.

I believe among other things relates to the emergency exits on the 737.

zeke wrote:

pune wrote:

I am failing to understand the bit shared above.

There is an old saying, better the devil you know.

The 737 built today is very different from the first 737. Whenever you build something from scratch it has unknown problems. Over time these problems should be resolved through in service experience, service difficulty report, or accidents.

Building on the baseline of that improved 737 means many problems have already been addressed, so you are starting from higher level of safety.

We see this in all sorts of engineering worldwide, you learn from your previous designs.

The flip side of a coin is that things are getting messy. In 50 years, there are no people in the project who took part in original development. And things which took a lot of effort in the original design may be taken for granted and broken by those who don't realize why things are done in a certain way.
Building up on old design has its limits, at some point development may enter the road paved with best intentions.

Phosphorus wrote:

Another factor is the ability, via grandfathering, to be subject to older (presumably less stringent) requirements in certification.

Sometimes quite a lot less stringent regulations.

A large portion of the fractional weight difference between B737 and A320 will be due to the applicable requirements differences for both airframes.

kalvado wrote:

zeke wrote:

pune wrote:

I am failing to understand the bit shared above.

There is an old saying, better the devil you know.

The 737 built today is very different from the first 737. Whenever you build something from scratch it has unknown problems. Over time these problems should be resolved through in service experience, service difficulty report, or accidents.

Building on the baseline of that improved 737 means many problems have already been addressed, so you are starting from higher level of safety.

We see this in all sorts of engineering worldwide, you learn from your previous designs.

The flip side of a coin is that things are getting messy. In 50 years, there are no people in the project who took part in original development. And things which took a lot of effort in the original design may be taken for granted and broken by those who don't realize why things are done in a certain way.
Building up on old design has its limits, at some point development may enter the road paved with best intentions.

That is why usually in FOSS, we have comments which developers use when programming so that later when they feel the need to revisit a piece of code or one which is not tested (corner-case scenario) they could come back, read the comments and then fix or better it. Of course, this is the 'ideal' situation, how many developers do that is altogether a different scenario although usually in FOSS, people do take care, especially if distributions distribute your product and you have actual users (casual or otherwise) who file tickets/bugs or/and ask questions.

Again, it's a different world altogether because unlike in aviation where lot of things are private, here things are mostly the other way around. The only private thing may be how the project, in this e.g. debian-project is running but even that for quite few years has been mostly in the open.

https://lists.debian.org/debian-project/

There are still some things which are 'private' and perhaps there is need for it, but the idea is to have more and more transparency.

Nick614 wrote:

This isn't a software thing, it is a type rating that continues to be the same so airlines dont have to do months of training per pilot for the newer model. A320 NEO, A330 NEO, E2 jet and 777X will also fall into this category.

I guess a discussion could be around the regulations on what or how many changes can be made to retain the similar type rating.

That discussion would be interesting and enlightening for sure. Maybe we can use that as the basis for this thread. I was under the impression that 777X will be a new 'type' . One basic question is how do regulators determine what is the type and when they say it's a new type ? Is it at design stage or at the manufacturing stage or at the testing stage ? Because if it's at the testing stage then too late in the game as far as economics is concerned, isn't it ?

TheWorm123 wrote:

I must say never did I think Linux and this forum would ever cross over

The new and improved 737-MAX powered by the Raspberry Pi 4 Model B.

Nick614 wrote:

This isn't a software thing, it is a type rating that continues to be the same so airlines dont have to do months of training per pilot for the newer model. A320 NEO, A330 NEO, E2 jet and 777X will also fall into this category.

I guess a discussion could be around the regulations on what or how many changes can be made to retain the similar type rating.

I wonder which is worse. Changing the type rating, training pilots and getting the MAX in the air, or allowing 400+ brand new aircraft to grace West Coast tarmacs for months on end? The former option may be a disadvantage vis-a-vis the A320NEO but it will at least get the planes in the air.

How serious is the issue of pilot re-training for airlines? Sure, it will add to costs but it must be cheaper than keeping multimillion dollar airframes on the ground?

I am going to ask an honest question. Who here has read the certification basis for the 737MAX and/or the A320 NEO? I have seen lots of comments on things being grandfathered on the MAX where those systems are clearly not. In many cases where the MAX, or NEO for that matter, is certified to an older versions of Part 25 or CS-25 there are additional items and equivalent level of safety findings. All derivative aircraft will use older certification bases for some aspects of their design. This will be because those systems/components are unchanged and the changes in the regulation are such that rectifying the unchanged bit is deemed not to produce a meaningful increase in safety.

Some other things to note:

• Most brand new aircraft will be certified using a basis that isn't fully up to the most recent version of the standards. This is because the standards change during the certification programme
• The areas where 737MAX is having issues with regards to MCAS are covered by much older versions of the standards, ones that not only the MAX is subject to, but also the NG and classic 737s. Some are covered for all 737s
• Did you know that all Boeing transport aircraft certified after 1990 do not directly comply with the FARs applicable at the time their certification programme was started. They all have higher service ceilings than allowed by the standard. Boeing has received and equivalent level of safety for all of these designs.
• Grandfathering is about Type Design (Certificate) not Type Rating. Obviously any aircraft with a different Type Certificate will have a different type rating. However, these may be common Type Ratings. Good examples of the are the 757/767, A330/A340, 777/787, and others. In some jurisdictions at least all the Airbus FBW aircraft have a common Type Rating.

TheWorm123 wrote:

I must say never did I think Linux and this forum would ever cross over

I love a good analogy, but I think your comment is pointing to the deeper truth that there is little crossover, for some pretty significant reasons:

a) software is relatively easy to create, even in the old days the tools were easy to get, and plenty of people are trained in their use, and lots of info is available on how to use them and how to write code

b) as the thread starter points out, plenty of people are willing to work on foss for free whereas we don't find people volunteering to work for boeing or airbus for free

c) free software has almost no regulatory concerns at all, commercial aviation is heavily regulated

Here's my best attempt at a software analogy.

Let's consider the early days of personal computing, when Apple and M$ were the dominant players. Let's presume there was a high barrier to entry for competitors (due to patents perhaps, or access to microprocessors, whatever) so no competitors such as FOSS could emerge. In this world let's presume Apple decided to jump from the Apple II to the Mac fully knowing they would take the big hit with regard to needing to retrain the customer base, needing new software, etc. Let's presume M$ was able to avoid doing so and keep cashing in on MSDOS because the high barrier to entry meant no third party could undermine their strategy to stay on MSDOS. This would allow them to continue to charge high prices without investing heavily in future development and without forcing users to need to be retrained. In turn the end users who may have been tempted to move to Mac see all their old software will still work and their old skills are still relevant so they decide to not switch to Mac. The kinds of things M$ would invest in are kludges such as extended/expanded memory and terminate/stay-resident code to keep the goose laying the golden eggs.

Some portray the world in a similar way: Boeing has used their duopoly status to avoid costly redevelopment of the 737 and has booked large profits while doing little development. I think it's a fair criticism. Others (the one the thread starter quoted, for instance) say the regulatory environment is the source of the high barriers to change and a more flexible environment would have allowed for much more incremental development of 737.

As a thought experiment, could Boeing retrofit FBW to 737 and retain a common type rating? On one hand, if they made the effort to keep the "look and feel" the same for the pilots one could imagine it could be the same type with a large differences training requirement. On the other hand, one could envision the systems differences and thus failure modes would be so different that one could not maintain a common type rating. Yet we know if a common type rating can't be retained there is no way to make the business case for such a retrofit work and the end result will be kludges such as MCAS.

There is a lot of value in what this site refers to as grandfathering, and a lot of it comes from the fact that the regulatory environment makes developing new stuff very expensive and of course people talented enough to do the work want to be paid well for all the responsibility they take on, so there is no low cost path for new development. The problem is that there is also a lot of corporate greed and we can see it in action via the recent Boeing message dump. This value means there's a bias towards reuse with minimal change, and a temptation to not really consider all the ramifications of a change because there is a false confidence built up and because management expects changes to be minimal. It's clear Boeing sales made the $1M per airplane guarantee of no sim training long before the engineers could work through the ramifications of such a commitment. One the course was set no one involved seemed to be willing to question it, rather we saw people admit they had to use "jedi mind tricks" to keep the illusion of minimal change intact.

DIJKKIJK wrote:

Nick614 wrote:

This isn't a software thing, it is a type rating that continues to be the same so airlines dont have to do months of training per pilot for the newer model. A320 NEO, A330 NEO, E2 jet and 777X will also fall into this category.

I guess a discussion could be around the regulations on what or how many changes can be made to retain the similar type rating.

I wonder which is worse. Changing the type rating, training pilots and getting the MAX in the air, or allowing 400+ brand new aircraft to grace West Coast tarmacs for months on end? The former option may be a disadvantage vis-a-vis the A320NEO but it will at least get the planes in the air.

How serious is the issue of pilot re-training for airlines? Sure, it will add to costs but it must be cheaper than keeping multimillion dollar airframes on the ground?

From what I could get sense of in the big viewtopic.php?f=3&t=1437867 it is not just the pilot re-training would be required but also other invasive changes. From the data dump at least to me, it seems they did lot of short-cuts in engineering and design because the management's idea was to have a competitive product to the A320 as fast as possible. The recent data dump is small and we don't really know how much and how many short-cuts were taken. Should the U.S. Congress broaden the scope, as a paying customer I would hope so but am realistic enough to know that big business, lobbying and how corruption works. If it would broaden the scope and whistleblowers are compensated (as their careers would be over for forever) it would lead to lot more things that we would know about that perhaps we don't know today. There is only one thing clear which we know about the Boeing situation is that we don't know a lot.

To come back to my original question and the reason for the thread. could somebody share how regulators decide or tell that this is a 'new type' and what all goes into that consideration ?

To share a somewhat different but perhaps in some ways a similar example, for e.g. in Windows 10 and perhaps even before if you replace the motherboard, windows 10 no longer recognizes the license and you have to reactivate it

https://support.microsoft.com/en-us/hel ... are-change

also happens in Debian as can be seen in -

https://unix.stackexchange.com/question ... otherboard

Looking forward to answers.

DIJKKIJK wrote:

I wonder which is worse. Changing the type rating, training pilots and getting the MAX in the air, or allowing 400+ brand new aircraft to grace West Coast tarmacs for months on end? The former option may be a disadvantage vis-a-vis the A320NEO but it will at least get the planes in the air.

How serious is the issue of pilot re-training for airlines? Sure, it will add to costs but it must be cheaper than keeping multimillion dollar airframes on the ground?

Sorry, but there is no need to change the type rating at this point. Boeing has already agreed to change its position on sim training, which was the real issue. They are also redoing the pilot training material and check lists. This is largely because the regulators found current airmen weren't able to reliably interpret what the airplane was telling them and what they needed to do next. This is also a big part of why the RTS date just slipped out another three months or so. MAX will have a common (airman) type rating with NG, but a much larger differences training syllabus than what Boeing originally proposed and FAA approved.

phollingsworth wrote:

I am going to ask an honest question. Who here has read the certification basis for the 737MAX and/or the A320 NEO? I have seen lots of comments on things being grandfathered on the MAX where those systems are clearly not. In many cases where the MAX, or NEO for that matter, is certified to an older versions of Part 25 or CS-25 there are additional items and equivalent level of safety findings. All derivative aircraft will use older certification bases for some aspects of their design. This will be because those systems/components are unchanged and the changes in the regulation are such that rectifying the unchanged bit is deemed not to produce a meaningful increase in safety.

Some other things to note:

• Most brand new aircraft will be certified using a basis that isn't fully up to the most recent version of the standards. This is because the standards change during the certification programme
• The areas where 737MAX is having issues with regards to MCAS are covered by much older versions of the standards, ones that not only the MAX is subject to, but also the NG and classic 737s. Some are covered for all 737s
• Did you know that all Boeing transport aircraft certified after 1990 do not directly comply with the FARs applicable at the time their certification programme was started. They all have higher service ceilings than allowed by the standard. Boeing has received and equivalent level of safety for all of these designs.
• Grandfathering is about Type Design (Certificate) not Type Rating. Obviously any aircraft with a different Type Certificate will have a different type rating. However, these may be common Type Ratings. Good examples of the are the 757/767, A330/A340, 777/787, and others. In some jurisdictions at least all the Airbus FBW aircraft have a common Type Rating.

Agreed, there is a poor understanding of "grandfathering" on this site.

As explained to me during training, type certificates are for aircraft, type ratings are for airmen.

Grandfathering applies to aircraft not airmen.

As you point out, "grandfathering" is not a free pass. The basis of certification is agreed upon between the manufacturer and the regulator, and in many cases additional work is needed to be performed to assure equivalent level of safety.

Unfortunately we can see in the case of MCAS, Boeing decided to position MCAS as a component of STS to avoid having to evaluate it as a new system and used weak if not negligent evaluation of its safety category to avoid costly analysis and testing. Had it been more concerned about safety and less concerned about cost it could very well have avoided the two tragic crashes.

Quick look. The Global 7500, in systems and flight control laws isn’t very different from an A220; but separate type ratings and separate TCDS. But, the G7500 is an amendment to the BD-700 (Global Type), even with a clean sheet airframe, engines and FBW. Go figure. For pilots each of the three are different type ratings and wholly different training programs.

The FAA, for any design, new or amended, will convene a Operational Evaluation Board with the FAA and OEM pilots to decide on the required training, what can be differences training or what requires entirely new rating, and what can be common across types. Once the OEB evaluates and makes its decisions, by tabletop and flight profiles (aircraft and/or sim), the OEM produces the training plan and submits it for approval by the regulator.

GF

pune wrote:

could somebody share how regulators decide or tell that this is a 'new type' and what all goes into that consideration ?

You need to narrow down if you are talking about:

a) aircraft, thus type certificate ( ref: https://en.wikipedia.org/wiki/Type_certificate )

and/or

b) airmen, thus type rating ( ref: https://en.wikipedia.org/wiki/Type_rating )

The references provide a lot of background..

With regard to aircraft not airmen, https://www.transportation.gov/testimon ... tification has testimony by the FAA Deputy Administrator to Congress describing the process of type certification:

The aircraft certification process has four stages: (1) certification basis; (2) planning and standards; (3) analysis and testing; and (4) final decision and certification of design. The FAA determines the level of involvement of the designees and the level of FAA participation needed based on many variables. These variables include the designee’s understanding of the compliance policy; consideration of any new and novel certification areas; or instances where adequate standards may not be in place. The work FAA delegates primarily relates to analysis and testing. About 94% of work in this area is delegated, and that work involves lower risk and routine items. The FAA does not delegate the other functions. The FAA determines the certification basis, identifies the standards, and makes all key and final decisions.

So, in short FAA decides if it can certify on the basis of amending an existing type certificate or a new type is needed, with FAA leading the process and delegating much of the analysis and testing to the 'designee'.

DIJKKIJK wrote:

...

I wonder which is worse. Changing the type rating, training pilots and getting the MAX in the air, or allowing 400+ brand new aircraft to grace West Coast tarmacs for months on end?
...

Priority is proving that MAX can be safely operating, period. If this cannot be positively proved, they stay on the ground, period. Until plane airworthiness and pilot training are BOTH assured, anything else is irresponsible.