diamondaviators.net

Done. See viewtopic.php?f=8&t=6646

My details for the list:
DA40 TDI
S/N: D4.338
D-EWDI

G1000 Software Upgrade - Petition List a/o November 2, 2018

OWNER / PILOT REGISTRATION MODEL
1. Doug Wall N149WT DA42
2. Colin N972RD DA42
3. David N16FA DA42
4. Emir SEMAD DA42
5. Keith EGNH DA42
6. Reinhard DEWDI DA40
7. Neema N112NG DA42
8. Karl DGGWU DA42
9. Hans D-GDON DA42
10. Hans D-EWDI DA40
11. Scott Sutton N189Y DA42

I think we have a wiki to put things in. Or we used to. You can add my contact information to the list. Diamond might want to reach out to some of the people to check in with them, who knows.

Also, I see Dave's point if the cost is really in the millions, as per the Garmin contact. That seems like an absolutely bananas figure to me and I'd love someone to break things down. They keep talking about needing to test the software in a plane. What is that regime like? If I were going through Kansas (30th coast-to-coast crossing was last week, so I am often going through Kansas), could I stop and help? If they put an "EXPERIMENTAL" sticker on my plane could they load the new software up and I'd fly it for them for twenty hours and they call it done? How does it work?

But millions... yeah, I can't see Diamond doing that. Yes, I think there's a *really* good reason for them to show support for the plane and to learn (as a new company, which they sort of are) how to properly support the fleet. And I think the owners could bear *some* of the cost (like I said, I'd be happy to pony up even a full AMU per year to stay current, I have to do that with my drafting software), but there aren't enough of us (200 in North America?) to be an actual sales target. BUT, learning what is required to keep the software current is key.

Since Diamond doesn't build this plane anymore, is there any reason for them to continue to hold the STC? Could they sell it to the DAA for a dollar?

All good thoughts Colin! Perhaps one of the forum moderators can help us out with this "Wiki" option. Although I'm not entirely clear what this might be as I'm relatively new to this forum. Yet ANYthing that would save hours of data entry would be fantastic - as would Diamond reaching out directly to owners! Perhaps even actively joining this forum, so as to engage with all of us directly . . .

Also, maybe see if Trek would be willing to meet with you on your next stop in Kansas. One would think he must have a management contact (or two) at Diamond. So far though we only have eleven legacy G1000 aircraft that have expressed an interest.

If feels like this is going to be more difficult than getting an act passed by Congress, yet I love all of your thinking "outside the box" ideas. Just let me know when you'd like to receive your "EXPERIMENTAL" stickers in the mail. . . .

AMU stands for? (Atomic Mass Unit???)

Add my name, too (Peter, N442Z, DA-40-180, 40.1124).

On the software development side, the expense is likely in the six (6) digits. Figure it this way (and remember: everything here is a guesstimate built on no personal knowledge whatsoever):

- Software development: the DA-4X series requires customization of existing software for engine and airframe parameters. Without meaning offense, there are few parts to our airplanes that haven't been performance modeled. Our airplane's sensors, performance, and other characteristics are all ordinary.

Garmin is a perfectly competent software developer. I am sure -- without having any knowledge other than what I see from the installation and maintenance manuals -- that their software is modular, object-oriented, asynchronous, parallelized, documented, and with well-known interfaces.

Further, we know that Garmin has developed all of the functionality -- GTX-345R support, AFCS support, new UI functionality for the NXi, Flightstream compatibility, instant holds, yadde yadde yadde -- in other instances for other airplanes. I have yet to see something major that only the Diamonds have ...

Given all of the above, the work required is merely to adapt / re-adapt / conform / integrate existing software into the package needed for our airplanes.

Let's assume 1,000 hours of work here (half a year of someone's time), though it is likely much less. At $100/hour, this is $100k.

- Testing: never to be underestimated, the testing is also one of incrementalism. (The old software joke was that the first 80% of the time was on software development and the next 80% was on testing; today, this is much better as a 100% of the time is spent developing and testing in tandem.)

Someone has to do the unit testing (of the modules), integration testing (of the system, likely through simulation as already found in the CD-ROM versions), and then testing in the airplane.

How many hours? Let's say 1,000 (another half year of someone's life). Unlikely, but it matches the other 80%. At $50/hour, this works out to be $50k.

- Regulation: someone has to deal with the FAA. I've never done it for a product, but have seen a lot of regulated products in aviation and elsewhere get built. Lawyers, engineers, and regulatory affairs folks. Fees, too.

How much? Let's say another 500 hours. Unlikely, but what the heck. At $150/hour, this works out to be another $75k.

My math is pretty rusty and everything above has many assumptions, but I think it comes to $225k.

Now let's take another old adage and take my engineer's estimate and multiply by Pi.

Now we're at $706k.

Assuming that 1,000 airplanes buy the update then we see a cost per plane of, well, $706.00. Add in the usual aviation multiplier for retail, insurance, etc. and the cost should be still be less than $2,500 per plane.

This is software only. The new boxes (GIA-64s, displays, etc.) are all priced to make a profit by Garmin (probaby many times their actual bill of materials).

There are also likely licensing issues beyond the economics above where other manufacturers have exclusives with Garmin on certain functionality, either explicitly or implicitly. I wouldn't be surprised if Diamond is disadvantaged by other manufacturers with more volume or value/unit, for example.

Thoughts?

Your rough estimates and logic seem reasonable to me Peter. Thanks for your thoughts, plus signing up!

12. Peter N442Z DA40 (yet a/o November 3, 2018 - 527 views)

Assuming that 1,000 airplanes buy the update

Unfortunately it’s much less, I would say that even 100 is quite optimistic.

It’s not easy to estimate number of man/days needed for development/testing of software product you’re not familiar with in details but having working product certified for the same airframe makes life much easier. However, any analysis will lead to the conclusion that I’ve already written: it’s not millions, it’s not even one million, it’s much less.

And regarding certification, manufacturers always stress it’s expensive and long process. It is but usually only for the first version and on the first airframe. Later on they reuse previous ones to get new ones or simply claim minor changes not requiring re-certification.

Just take a look at DA42 - at EASA type certificate there are 5 types: DA42, DA42M, DA42-NG, DA42M-NG and DA62.

https://www.easa.europa.eu/sites/defaul ... e%2041.pdf

Looking what’s inhereted and shared you can get rough picture of level of work needed for certifying some system that you already have certification for in the same airframe.

BTW Does anyone know what this brought and did anyone install it?

http://support.diamond-air.at/fileadmin ... ftware.pdf