From: NASA Blogs
Posted: Friday, August 21, 2009
Alan Hockstein was the man most feared by pilot astronauts. Well, except maybe for George Abbey. Let me explain why.
The shuttle is the world's largest glider. The pilot has one and only one chance to make a landing; there is no "go-around" capability. Obviously, good piloting techniques are studied exhaustively. Much analysis and simulation has been completed to maximize the chance for a successful landing.
Alan was the senior landing analyst. That means he studied more and worked harder than anyone to understand how the shuttle flies - especially in the final approach and landing phase. One part of Alan's job was to analyze the telemetry from each shuttle landing and see how that compared to the "ideal" landing. So in a quiet office environment over a couple of weeks, Alan and his team would look at each telemetry point, every sample (up to 125 per second for some parameters) and compute how each one affected the landing.
Every shuttle commander dreaded the day of the Entry, Descent, and Landing Debriefing. Standing in front of a projection screen filled with data curves in the presence of a room full of folks, Alan would ask the commander something like: "why did you deflect the hand controller here" pointing at a squiggle on the screen. "That input caused a deviation of 12 feet high above the flight path which correlates to a 273 foot miss distance at the touchdown point." The commander would squirm in his seat and say "we had a wind gust" or some such. Alan would point to another squiggly line on the plot and say "the accelerometer data doesn't show a wind gust at that point." The poor pilot would then have to come up with some other lame excuse: "the visual scene was obscured by some wispy clouds." Alan would pull out the meteorological report "the lowest observed clouds were at 25,000 feet" And so it would go. Excruciating for the veteran test pilots who pride themselves on their steely nerved stick and rudder reactions.
Why did we go through this ritual? One reason only: to learn what we could about flying techniques, how they affected the landing, what might work better. All of this so that the next pilot would have a better idea of how to maximize the chance "for a happy outcome."
At our Flight Techniques meetings, Alan was a frequent presenter showing what had been learned, advising of the best techniques. At one period we experienced a number of landings that were shorter than desirable - still on the runway, but consistently closer to the threshold than comfortable. Alan analyzed hundreds of combinations of factors over a several dozen landings looking for correlations. Nothing seemed to correlate, except one: "If you cross the threshold low, you are likely to touch down short."
Now that may seem obvious in retrospect. If a glider comes in low, any pilot would intuitively expect a short touchdown. But it was only obvious in retrospect. And any number of other correlations that common sense might have suggested were simply not borne out by the data. So we called this "Hockstein's Law": If you cross the threshold low, you will touch down short. The entire community worked very hard with the pilots to improve techniques to be higher at threshold crossing and thereby the incidence of short touchdowns was significantly reduced. Well, that is the very short summary anyway.
Nowadays, I don't spend my time studying shuttle landings like I used to. Recently I've been a data gatherer and logistics helper to the Augustine Committee. That group has been getting a lot of data and, among other things, looking at the cost estimates of various options for space flight. I'm not well suited to work in that ethereal regime; nuts and bolts are more my specialty. But it occurs to me that we need an Alan Hockstein to look at project development budgets for clues of how to improve the performance of future work.
Somebody who will look at each data point in depth, spend the time to think about it, calculate the consequences of each movement, and then provide those of us who may have to execute a project in the future with some guidelines that might lead to a better likelihood of a "happy outcome".
Some of my experience suggests possible correlations between different events and poor program performance. For example, continuing resolutions on the budget cause disruptions and delay planned activities. It would seem that there might be a high correlation between lack of a firm budget (e.g., a continuing resolution) and poor program performance. Then again, Norm Augustine himself kept saying that the secret to successful project management is reserves. Perhaps the congressional prohibition against budgeting reserves for projects plays a role in poor program performance. Then there is something called a "rescission." I never knew what a rescission was until I got into program management. A rescission basically prevents a program from spending all the money budgeted for it. I'm no analyst but it may be that rescissions play a role in poor program performance.
All of those things are just guesses on my part. I'm no analyst. But it seems like a good study if we want to have successful projects in the future.
Now, where is Alan when we need him?
// end //