"Orbits" Part 2

In my previous post I described some of the technical hurdles we encountered during the development and certification of the Garmin GSR56 Iridium Transceiver. The story continues with the new Garmin GMA350 audio panel. We had been evaluating some prototypes of this new audio panel and wanted to put it into production, but we also wanted to be able to offer a GSR56 Iridium transciever retrofit kit for existing users to install in their airplanes that have the existing GMA347 audio panel. This meant we had to test the Iridium system with both the existing GMA347 audio panel and the new GMA350 audio panel.


What this means for us is that we had to install and test this in an airplane twice. The wiring for the two audio panels is completely different which complicates things. We had to take one of our experimental airplanes which has the GMA347 audio panel and  install the Iridium box and associated structural modifications, hardware and new antenna. We then did all the testing with that version of the installation. Then we took out the old audio panel, rewired the airplane for the new audio panel and repeated all the testing for that installation.

Software is another area where people tend to think it is an easy change, but we spend a lot of time doing enough testing, in many different conditions and scenarios, to ensure that there are no bugs in the software. This involves not only on-aircraft testing, but also installing it in a "bench" installation in our lab. We can simulate a lot of different warnings and other scenarios on the bench before actually doing the flight testing.


But real world results come from the extensive ground and flight testing we do. Every flight test program ends up with some interesting stories and this project is no different. One interesting bug we discovered was that a phone call would not terminate when you clicked hang-up on the aircraft but the guy on the other end of the phone didn't hang up (like when dialing into your voice mail.) This has since been fixed.

Another instance occurred when we were proofing out the voice service. From the ground we called the test aircraft and the guy talking from the plane then sat through a 45 minute teleconference. What we didn't know was that when we dialed the aircraft from our land line it was being considered an international call by our company phone service which racked up quite a charge. Oops.


So after all this testing we finally got a solution that we were happy with and began the process of creating all the various drawings (wire harnesses, installations, structural changes etc), and summarized all of our testing through plans and their test reports, along with a couple of certification reports and submitted that to the FAA.

While that was going on, we sent an airplane down the production line to act as a "pilot build" aircraft. We use this aircraft to proof out our drawings, any new processes, parts and installations. It also provides a chance to give the factory technicians training if necessary. We then feed back any information we learn to correct drawings or re-orient the line flow prior to the need to build these airplanes at rate.

With all of that complete, and final approval from the FAA we are able to confidently move forward with production knowing that there will be a few minor changes needed but overall production will proceed smoothly. http://img.ezinemark.com/imagemanager2/files/30004252/2011/11/E10.jpg