|Orville demonstrating the flyer|
to the U.S. Army, Fort Myer, Virginia
September 1908. Photo: by C.H. Claudy.
Why did they succeed where others failed?
Not by luck but by their methods. Rather than testing entire designs in total, as did many others, they developed and tested a component at a time following a repeating cycle of research, calculation, experimentation, and refinement.
Their successful powered flight and total control of the aircraft wasn't a surprise to them. It was the culmination of four years of intense, methodical, purposeful, and driven research and experimentation.
By the time they flew at Kitty Hawk under power, they had
- tested many airplane control designs using kites,
- created their own wind tunnel and balance for more sophisticated testing,
- evaluated more than 200 different wing and airfoil designs,
- piloted their own glider aircraft more than 1000 times
- proven wrong, through experimentation, the established coefficient used to calculate lift, and
- through wind tunnel testing, developed propellers that modern wood props can barely outperform
Breaking down complex problems into smaller problems is a recipe for success. So is designing and experimenting with individual components under controlled conditions prior to integrating the components into the finished design.
The biggest cause of my 2011 Sparkfun AVC failure wasn't due to its myriad flaky sensors. The sensor problems were merely a symptom. I didn't experiment enough with each sensor component.
For example, I encountered magnetometer (compass) errors. Was it because of accelerometer-based tilt-compensation distorted by linear acceleration? Distortion from nearby metal objects? Or DC motor magnetic fields?
I had no idea. Because I didn't experiment individually with the magnetometer sensors. I just slapped them on the robot and drove the whole thing around haphazardly. I tested the entire system, not its components.
So it's time to start over, experimenting in a more controlled fashion with each sensor, magnetometer first.