Colyn,
No offense taken. I've given this a great deal of thought (and have already put considerable time and money to fix the autopilot) already.
"
circumstances create workload that exceeds pilot capability" is spot on. This last flight was an absolute cakewalk compared to one I had last year which really did take me to my workload limit (see this forum post for info, http://forums.pilotedge.net/viewtopic.php?f=18&t=1681, but it's a flight that' I've posted on the LML before, including the MP3 recording). It also became a scenario for the PilotWorkShops IFR Mastery Series.
I do plan on upgrading my weather capability (it's a small investment to get a Stratus, which will provide ADS-B weather) and if circumstances permit, I would like to have the autopilot. However, down low, in the soup and the bumps, I am far more comfortable hand flying than with an autopilot for three reasons:
1) I worry about what happens if the autopilot doesn't do a perfect job and I'm off doing something else because George is flying. In other words, it's going to be tough to relinquish control and focus on other tasks without allowing myself to become to engrossed in those other tasks
2) if there is a performance issue, I become aware of it within a couple of seconds when hand flying. With George flying, and a reduced scan, I might not find out about it until the onset of a stall (extreme case, but you get the idea).
3) this is a point I'm adding to the list having finished writing the rest of my piece. I tried to reconcile the fact that I do coupled approaches in the sim, but not in real life. When I really think hard about, I simply don't trust the a/p configuration I have right now. If I did have an a/p that I could trust, then, during an approach in smooth conditions, I could see myself relinquishing control, and that being a good thing. It would affect my scan..I wouldn't use the workload reduction to do other things, I would simply maintain the scan throughout the approach, and little else. Overall, I agree, it would reduce overall workload. Anyway, back to the original message,...
The few times I have engaged the A/P on long XC legs, I find myself less engaged than when I'm hand flying. Case in point, just west of Washington DC, I routinely get soft, subtle 300-500fpm updrafts and downdrafts. I know about them right away when flying by hand, and it's valuable information.
My strategy has been to maximize my workload capability through simulation with online ATC, and to ensure that regardless of how busy it's getting, to never stop flying the plane. On that 'triple diversion' flight posted above, the reason it took me so long to brief the approach at TEB was largely because I was allocating a healthy amount of time and brain power to flying the plane, managing fuel, etc. I chipped away at the task of pulling up the approach plate for the ILS RWY 19 at TEB, along with getting that confounded ATIS (don't even get me started on what a chore that is when New York Approach is belting along at full throttle because they're running combined sectors at night).
Adding a second pilot is a complete non-starter. I read that as a proposed solution quite often and just shake my head. I can't think of anything less practical for these long trips, and as such, making it a goal, or even coming to rely on it as a safety feature is a non-starter for me. The upgraded weather capability, agreed, I should and will do it. The auto-pilot...I plan on doing it when I can, but probably still won't use it in the circumstances that others would elect to us it. I'm more likely to use it in VMC on a long leg to enjoy the view a bit more.
Another way to look at the autopilot issue is that not having one that I can bet my life on has lead me to become completely self-reliant on hand flying skills...and I am completely comfortable with that at this point. Take those who have the luxury of having bullet-proof auto pilots. Are they routinely practicing hand flying their approaches in a variety of weather conditions? It's hard to generalize, of course, but of the population of pilots, I have to imagine there are some that have become so reliant upon the auto pilot that, practically speaking, they're really not comfortable doing it by hand because the perceived workload is just too high, and their proficiency at doing it by hand just might not be there (I do tons of coupled approaches in my sim, btw, and the thought process is a completely different animal). Hence, the autopilot becomes a "must have" before launching IFR. That's not an attack on their potential abilities, it's just a theory that I have about where their proficiency might be (again, for SOME pilots, not all, there will always be exceptions).
I hand fly all approaches in real world (for now). I hand fly maybe 50% of my approaches in the sim, and use varying levels of automation for the remaining 50% down to various points along the approach (I try to expose myself to as many circumstances as possible). Please know that I'm not hand flying them to test myself or prove any points, I am just comfortable relying on that skill set AND I don't have a real alternative right now, nor is it an urgent priority to change that (but it IS on the list).