According to the President of the EAA, the hotspots are first flights, 50 hours, anything unduly hot (like Lanceairs) and the turn from base to final for LSA.
The question is, what can the FAA do, regulation-wise, to improve the accident rate in these categories?
I have direct experience here. I was a member of an FAA Flight Safety Board last year, examining how the EX-AB safety record could be improved. There were about twenty-five members, mostly FAA, with representatives from the EAA, AOPA, and some of the builder's groups. We met four times, mostly in Kansas City.
The main problem is that there were few key aspects that could be pointed to, as a cause. And the few that were found had no real solutions.
"More training," some of the FAA participants declared. "We should require that homebuilt owners get ten hours of instruction in type before carrying passengers."
The trouble is, how does the builder of a Smythe Sidewinder find an airplane to take dual in? How does he "rent" the airplane when the FAA doesn't allow that sort of use? Where does he find a qualified CFI?
Even where potential solutions were found, the benefits just weren't there. For instance, one of the concerns was the accident rate among recent purchasers of flying homebuilts. A couple of the senior FAA types were advocating requiring that newly-purchased homebuilts be placed into the phase 1 flight test limitations again (e.g., no passengers) for ten to twenty hours.
Nice idea, but such a restriction isn't really justified by the statistics. Over the ten-year period covered by my own analysis, there were about 750 accidents involving purchased homebuilts. Of those where the pilot had ten hours or less in the aircraft, only four passengers were killed. So such a rule (if the homebuilt world strictly abided by it) would save just one life every two years.
So, what could the FAA do?
Require check-outs before carrying passengers? Per above, they'd have to devise some way to do this for less-common types.
Increase the test period for new homebuilts? The accident rate drops to "normal" at the 40-hour point as it is (albeit it jumps up a bit from 40-60 hours as planes start flying cross-country--VFR to IFR, running out of fuel, etc.--and mechanical problems start arising from use).
Require annual inspections be performed by A&Ps? Many A&Ps won't touch homebuilts. In any case, my statistics show maintenance errors cause less than 5% of homebuilt accidents. Not much leverage, there.
About 55% of homebuilt accidents are due to pilot error (including judgment errors). The percentage is even higher if NTSB "Probable Cause" is used (the NTSB blames the pilot if the investigator feels a forced landing could have been safely made after an engine failure, where my records reflect the cause as mechanical error). The next leading cause..."Undetermined Engine Failure"... is about 8.5% of the total.
So the biggest bang for the buck, here, is pilot training. Which, considering the wide nature of the homebuilt fleet, is the most difficult to regulate.
The FAA would certainly be capable of instituting a shotgun policy with no real basis in reality. But the aviation groups will oppose, and those members of Congress on our side will get involved. Without some sort of statistical justification, FAA officials know they'll come out looking pretty bad.
Van is right to be concerned. One out of every six homebuilts on the rolls is an RV; 30%-40% of new homebuilts every year are RVs. This means the investigators are seeing a lot of RVs, and the sins of the homebuilt fleet are likely to unfairly attached to the RV line.
Ron Wanttaja