I want to comment on your Safety column in the June 2005 issue (page 65) titled "Perils of Single-Point Failures." Let me begin by saying that occasionally, when reading your articles, I have not always agreed but have always enjoyed your "take" on the subject or issue.
I was inspired by your article on the extended consequence of a single-point failure of an inverter in a de Havilland DHC-8. I hope that technicians, engineers and, particularly, managers in the aviation industry read your column and consider your conclusion.
I once was an FAA designated engineering representative (DER) and have worked on aircraft certification and design for the last 10 years at major aircraft companies that build business jets. I now work in the Aircraft Certification Office.
From this perspective, I voice my concern. Should this incident on which you commented have occurred? I say, no, believing that a properly completed faults mode and effects criticality analysis (FMECA) would have addressed this problem in the design state. I think most people in the industry would concur with your assessment.
However, while many aircraft manufacturers give more attention to the safety analysis process, there still are those that question the process's value. Some simply ignore it.
We at FAA have made great strides with many companies in giving safety analysis a high priority up-front in the design process. Too often in the past this meant simply checking a box long after the aircraft design was complete. The result: last-minute fixes, increased costs and ultimately delays in schedule. Some would contend that safety took a back seat to cost and schedule.
Today FAA is delegating more responsibility to many of the companies designing and building the aircraft for tomorrow. Much of this delegation involves making the critical design decisions pertinent to safety. My hope is that small and large manufacturers, alike, do not choose cost and schedule over safety, but rather keep examples such as yours in their rear-view mirrors.
I've just read your useful and informative Perspectives column in the July issue (page 46), titled "Software Specs: The Next Step" and written by David Watrous of RTCA. While reading I noticed one inaccuracy. The article says "Airborne system software was introduced in the 1970s." But actually it happened much earlier. From my recollection the first airborne inertial navigation systems were introduced in the late 1950s, and flight management systems with complex software were available in the 1960s. As to military avionics, by the middle of the 1960s software as complex as the real-time Kalman filter already was implemented, in the C-5A Galaxy transport aircraft.
This small inaccuracy could lead to non-adequate understanding of avionics achievements of the past.
Navigation Analyst, Team Leader
Richmond Hill, Ontario
I enjoyed your article titled "LPV or RNP?" in the August issue (page 18). There is, however, one small item in the next-to-the- last paragraph, which may confuse readers. It reads: "Airbus...plans to offer WAAS/SBAS as a customer option."
Airbus does not plan to offer a WAAS/SBAS option but will continue to assess new functional capabilities, such as Galileo and GPS L5, in the context of adding value to customers.
Director, Air Traffic Systems