Rather than inject some of my personal history into discussion of another man's fatal accident, I set up this separate thread.

Quote Originally Posted by 1600vw
Ron let me be the first to say thank you for up dating all of us on this. What I learned from not only this but just flying Experimental airplanes. Do not treat one as a certified airplane. If you do it can and will bite you.

Also Ron. Great read in this months EAA mag. You do some awesome work. Do you go out to an accident or do you report or write about the cause and effect? Also how did you get involved in writing these reports? I find this very interesting. What you do that is.
Tony
I got interested in analyzing homebuilt accidents in probably the second-worst way: Standing in front of a TV camera with a reporter who had preconceived notions about how dangerous those home-made airplanes were.

(the worst way involves the wreckage of one's OWN airplane....)

Anyway, it was a homebuilt weekend at Seattle's Museum of Flight, and one of the local TV stations was pushing me to talk about homebuilt aircraft accidents. I mentally dredged up some of the statistics I'd read in various magazines and ummed and awwwwed my way through it.

(not all the coverage was negative. See: http://www.bowersflybaby.com/stories/PI_pic.jpg)

Anyway, after I got home, I got to thinking about the statistics I'd given the reporter. These had all be remembered from various magazine articles, but I really didn't know their history or accuracy. I'm an analyst by heart (a lot of my day job involved statistical analysis of satellite-target characteristics), and when I found out the NTSB database was downloadable, I extracted one year's database and set to work.

My initial goal was pretty simple: I wanted to find any recurring types of mechanical issues that affected given homebuilt types, to help find trends. That initial pass broke that year's accidents down to 30 mechanical-related issues, five pilot-judgement ones, and *one* for pilot error.

I gave a program at Chapter 26 on my results, but really wasn't too happy. A typical year sees ~200 homebuilt accidents, spread across hundreds of different type of aircraft. Not really enough, for one particular type, to spot any trends. Also, during the presentation I was asked for more detail about pilot error, so I expanded it into three sub-categories... Loss of Control (Stall), Loss of Control (Winds), and Loss of Control (Other).

I downloaded previous years (back to 1997). That resulted in more accident categories, as the "Other" category saw more and more accidents of a given type and I was forced to split them out. The next year, I gave a chapter presentation based on three years' statistics. A few years later, I had a full ten years of data, and started writing magazine articles.

My database has expanded to 65 accident categories, of which nine are pilot-error related ones.

I'm a data analyst; I've never been involved in the investigation of any particular accident. Oddly enough, though...I teach the investigators. The Department of Transportation's Traffic Safety Institute offers a number of courses for FAA and NTSB personnel. One involves investigating homebuilt accidents. It's a three-day course, with presentations by various specialists in a number of fields. EAA hosts this course out of Oshkosh, and when they set it up, Joe Norris (now with Sonex) taught a segment on homebuilt aircraft construction techniques. Joe had to bow out a few years ago, so I took over. It's even a paying gig, so I'm a government consultant!

With my retirement earlier this month, I'm looking into doing some in-depth analyses. I'm hoping to gain some insight into the impact of Sport Pilot on medical-related accidents. The NTSB recently released two accident reports involving pilots flying under SP who suffered major coronary problems.

http://www.avweb.com/avwebflash/news...-228648-1.html

I'm hoping to track the occurrence of such events in relation to the pilots' medical certification status over 18 years or so.

Undoubtedly will post the results here, but don't anticipate starting for several months.

Ron Wanttaja