Feedback needed: Should I clean up the format? Should I hyperlink or just leave it as plain text? What else do you like or not like about the layout? Send a tweet to @FlightTestFact with your thoughts.

What is your favorite airplane and when was its first flight? Send a tweet to @FlightTestFact to see it featured.  Search #flighttest on twitter for more.

11/1/1942 First flight #flighttest of the Westland Welkin. History of development of Welkin and related variants

11/2 First flight #flighttest of Lockheed XH-51 (1962) & CL-475 (1959), both rigid rotor experimental helicopters

11/3/1955 First flight #flighttest of the Martin B-57, eventually @NASAJohnson WB-57, high-alt research

11/4/1968 First flight #flighttest of the Aero Vodochody L-39 Albatros common USAF #TPS curriculum aircraft

11/5/2005 First flight #flighttest of the LS11 open class sailplane, in development at University of Cologne #video

11/6/1957 First flight #flighttest of the Fairey Rotodyne, compound gyroplane #video

11/7/1976 First flight #flighttest of the @Dassault_OnAir Falcon 50, first transatlantic biz jet

11/8/1970 First flight #flighttest of the P-300 Equator, by G Poschel. #Story: #photo

11/9/1978 First flight #flighttest of the McDonnell Douglas YAV-8B Harrier II via

11/10/1985 First flight #flighttest of the OK-GLI, analog Buran test bed, the Soviet space shuttle. #video

11/11/1956 First flight #flighttest of the Convair B-58 Hustler #video

11/12/1959 First flight #flighttest of the Avro VZ-9 Avrocar saucer #video via @afmuseum

11/13/1947 First flight #flighttest of Armstrong Whitworth AW-54 flying wing #video #history

11/14/1943 First flight #flighttest of the Lockheed XP-49, evolution of the P-38 #aviation #history

11/15/1929 First flight #flighttest of the DB-70, carried passengers in airfoil section

11/16/1962 first flight #flighttest of the XAZ-1 Marvelette, MSU test bed for ducted fan & boundary layer control

11/17/1956 First flight #flighttest of the @Dassault_onAir Mirage III #avgeek
11/18/1955 First powered flight #flighttest of the Bell X-2 via @NASAArmstrong

11/19/1948 First flight #flighttest of the Hawker P.1052 experimental aircraft

11/20/1912 First flight #flighttest of the Avro Burga, lateral flight control test bed

11/21/1952 first flight #flighttest of the Percival Pembroke P.66

11/22/1954 first flight #flighttest of the HZ-1 Aerocycle / YHO-2 / DH-4 Heli-vector

11/23/1942 First flight #flighttest of the Vought V-173 Flying Pancake

11/24/1959 First flight #flighttest of the Hiller X-18 experimental cargo-transport tilt wing/tiltrotor

11/25/1964 First flight #flighttest of the Convair Model 48 #video #news via @flightglobal

11/26/1991 First flight #flighttest of the Grob GF 200 pusher business aircraft

11/27/2012 First flight #flighttest of the @EmbraerSA Legacy 500, their first with fly by wire #avgeek

11/28/2008 First flight #flighttest of the COMAC ARJ21 Xianfeng via #COMAC The world awaits its delivery.

11/29/1974 First flight #flighttest of Boeing Vertol YUH-61 in the USA Tactical Transport Aircraft System competition

11/30/1979 First flight #flighttest of the Piper PA-46 Malibu via @piperaircraft

Check out @FlightTestFact on Twitter, Pinterest, or Google+ for more flight test safety references, videos, and information daily.

Other first flight indices: First - January | Previous – October

The dictionary defines sift as a verb meaning to examine (something) thoroughly so as to isolate that which is most important or useful.

How do we know what is most important or useful? I don’t know, but I do believe that applied tools of mathematics and statistics are in the top strata. In this environment, the demands on the flight test professional are ever increasing–added administrative responsibility with experience, schedule pressure, or even design complexity, just to name a few. The last thing I want to do is add something, but illustrating the wide applicability and high frequency of statistics, in particular, may help to convince us of its usefulness. However, statistics often defies intuition, and for that reason, I feel a collection of examples relevant to the field of flight test is a noble purpose. Mathematician Paul Halmost agrees:

A good stack of examples, as large as possible, is indispensable for a thorough understanding of any concept, and when I want to learn something new, I make it my first job to build one.

SIFT (Statistics in Flight Test)

This column, Statistics in Flight Test, will accomplish both of the aforementioned objectives: illustrate the frequency of statistics in flight test examples and build a collection of cases for us to study the importance of this applied tool of mathematics.

Here is the first example: RNP. A ninety-five percent confidence interval defines the width of an RNP position estimate as seen in the FAA figure below. As depicted by the yellow lines, RNP 1 is one mile from the yellow line to the center. (We won’t define confidence intervals here. That will have to wait for another post, however, due to their subtlety and intricacies.)

Flight test is the place where validation of the accuracy of navigation systems occurs by comparison to a truth position source.



Previous: 2 Kinds of Data | ATOMs Index

Follow @FlightTestFact on twitter for more examples of Statistics in Flight Test, or visit Rose Petal Press to download the authoritative flight test handbook and other rocket fuel for your brain.

autumn tree Winston-Salem North CarolinaAs I write this, it’s mid October, and I’ve just returned to the North Carolina Piedmont region from a five day trip to Arizona. When I came home Saturday it was as if the colors had suddenly changed–autumn had erupted into the brilliant colors that I love.

The truth is, they had not suddenly changed. Instead, it was because I had some temporal distance from the data that I noticed the difference so vividly. You’ve probably experienced something similar.

This phenomenon illustrates some important analytical principles in the following three broad categories: Forecasting, Strategic vs. Tactical, and Continuous Improvement.

I know the leaves will change color. I know that a tree will grow taller and thicker as it follows a seasonal cycle–autumn, dormancy in the winter, and growth in the spring and summer. This happens every year. We can forecast, predict with certainty, that this cycle will continue.

As pilots, we chart a course. This is also a kind of prediction or forecasting. We make this plan to a certain level of detail. We don’t do it to the finest resolution possible. That would require an inordinate amount of time, an investment not worthy of its diminishing returns. Additionally, there is a certain level of detail beyond which prediction is impossible. I can estimate where I will level off at my cruise altitude, but ATC constraints and delays make it unlikely that my estimate will be accurate.  Nor can I predict an exact ground speed, because I don’t know (exactly) the winds aloft.

There are limits to our ability to forecast, to plan, to predict.

We also have certain strengths and weaknesses when it comes to executing these plans. Our plans are strategic, and the actions required to implement them are tactical.

Strategic vs. Tactical
We are better at tactical. At doing the next step. Most of us aren’t good at strategic. Most of us either forget to think about or can’t intuitively grasp the impact of skipping an investment in our retirement account this month. We can’t visualize the cost of investing 7% instead of 15% into an IRA. (Dave Ramsey can and does here.) The same goes for our eating (dieting) and exercise habits. Flight test and leadership and data analysis in general are no exception.

It’s a paradox: We take small steps, but deviations–those times and place where we stray from the plan–are more apparent over longer range.  For example, where do you look when you drive a car, ten feet in front of your car or hundreds of feet? A gun sight, just a fraction of a centimeter low, could result in a shot that misses the target completely. This is apparent in attitude flying, where the horizon is infinitely far away, and very small differences in attitude are obvious, maybe even more obvious than on the attitude indicator.

In the fall, we fail to notice the minute changes in foliage. When we go to a different state or come home after a five day trip, however, we suddenly notice the distance.

I see the same thing happening on my drive to work.  I see green tree after green tree lining the side of the road, until I reach the crest of a hill that overlooks a valley covered in trees and catch a glimpse of an entire landscape in the same space of time that I normally observe individual trees. Then I notice Autumn’s beautiful color palette, the onset of the trees’ retreat into barren winter.  It’s interesting to note that Mother Nature doesn’t believe in continuous improvement.

Continuous Improvement
The trees go through seasons of growth and then dormancy. Their “improvement” or growth is better measured on a scale of years than on a scale of days and even months.

There is a phenomenon in human performance called regression to the mean. If someone plays well in a sports match, then it is likely the next game will result in a performance that is closer to the average.  If someone does poorly, perhaps playing an instrument in a recital, then it is more likely that their performance will improve the next time, thus returning (regressing) to the mean performance level. Human performance is cyclic, much like the trees, in that there are highs and lows.

This suggests that continuous improvement, in its purest application, is not possible. Certainly, the strength of our computation and data collection tools confounds our ability to achieve truly continuous improvement. We can detect variations in performance with incredibly precise measurements more quickly than we could in the past.

However, if we benchmark the performance on a monthly or yearly basis, we begin to see an upward trendline. We begin to see regular improvement.  There is some scale that shows “continuous” improvement, just as there is some amount of time that I need to stay away to see a sudden change in fall colors and some step size beyond which tactical increments give way to strategic achievement.



Related posts
Straight or Jagged Line | Previous: 2 Kinds of Data | ATOMs Index