CEV Abort System / NASA looks at Onboard Abort Executive

[jamie_young]

I have no idea about programing, but Texas Tech are working with the Johnson Space Center on the programing language and software that can think for itself and decide on the best course of action if the CEV has problems, including if it needs to do a Crew Escape abort. If it's good they might use versions in other parts of the CEV.<br /><br />http://www.nasaspaceflight.com/content/?id=4153<br /><br />But just now a programer expert posted this on the thread that goes with the story. This is just part of his critism.<br /><br />"Maybe I'm too old-school, but this doesn't strike me as good news. It sounds like somebody in NASA saw a "shiny object" and made a decision based on the glitz and glamor of these "natural language" tools. The system they have presented here is, in my opinion, overblown. It is based on specious theory and an untried language that, if carried to completion, will have as its first implementation our CEV abort system."<br /><br />Any programers on this site? It sounded very cool until that post above.

 

[najab]

As long as the Commander still has his rotary switch/pushbutton or abort T-handle, and as long as that has the final say, it's a good thing.<br /><br />If the Commander's switch is nothing but a "high level algorithmic input" (or some similar such wording), it's a <b />bad</b> thing.

 

[jamie_young]

But what if there is no time? Will the commander always have enough time to react to a serious problem to abort with the crew escape? I do understand that it is not maybe the best idea to leave it for a computer to decide, but something like Challenger where by the time the words uh oh were heard, it would have been too late for this crew escape in a CEV situation?

 

[arconin]

An automated system is desireable, so long as it actually works, but no commander should be beholden to a chunk of silicon, the commander needs the ability to abort if he feels it is necessary, a button or handle that will cause an abort regardless of what the computer decides.

 

[CalliArcale]

<blockquote><font class="small">In reply to:</font><hr /><p>But what if there is no time? Will the commander always have enough time to react to a serious problem to abort with the crew escape? I do understand that it is not maybe the best idea to leave it for a computer to decide, but something like Challenger where by the time the words uh oh were heard, it would have been too late for this crew escape in a CEV situation?<p><hr /></p></p></blockquote><br /><br />Well, first of all, the way Shuttle is engineered, there's no way they could've escaped. From liftoff to SRB sep, there is no abort. Maybe ejection seats could've given them a chance, but it'd still be awfully risky -- they were already going past the speed of sound. I guess it would depend on how early (and how accurately) you could get the system to detect the undesirable situation and trigger the abort. After all, you don't want a system that aborts too often; aborts are themselves risky, not to mention expensive.<br /><br />But I understand what you're asking. I think that the Commander and probably also the Pilot as a backup should have the capability to trigger an abort at will. People are smarter than computers; slower, yes, and not neccesarily capable of processing as much information, but we can still pick up on things that a computer would not only not recognize as a problem, but which the computer's designers may never have forseen. So an automatic abort wouldn't be bad, but there must also be a means of manual abort.<br /><br />Random factoid: the Soviets tended to take a different philosophy to manned spacecraft than the Americans. The American philosophy is typified by Von Braun's statement that a human being is the most sophisticated computer that can be built by unskilled labor; we put the people in charge because we want that capability in our spacecraft, even if many of the spacecraft's functions will still have to be delegated to a computer. The Soviets, on the other hand, tended to distrust huma <div class="Discussion_UserSignature"> <p> </p><p><font color="#666699"><em>"People assume that time is a strict progression of cause to effect, but actually from a non-linear, non-subjective viewpoint it's more like a big ball of wibbly wobbly . . . timey wimey . . . stuff."</em>  -- The Tenth Doctor, "Blink"</font></p> </div>

 

[josh_simonson]

To most software guys, a customized language might sound like a liability, but software meerely tells hardware what to do, so custom hardware requires custom software. Many complex pieces of hardware use their own programming language, which may look something like pigeon assembly. It's a lot of effort to set up, but it avoids the inefficiency and latitude for error involved in going up and then down a few levels of abstraction with every command.

 

[nolirogari]

Intresting that I've not seen mention of one key event here... Gemini 6.

 

[subzero788]

<blockquote><font class="small">In reply to:</font><hr /><p>Intresting that I've not seen mention of one key event here... Gemini 6.<p><hr /></p></p></blockquote><br /><br />Thanks for the reminder of Gemini 6, nolirogari. I'll reiterate the story here for those who don't know the story of Gemini 6’s failed launch. I’ve quoted Michael Collin’s excellent book, <i>Liftoff</i>. The crew were Wally Schirra (Commander) and Tom Stafford (Co-pilot):<br /><br />"Their countdown was trouble free, and the two Titan main engines ignited on schedule. At the same instant, however, their was an abrupt silence. The engines had stopped! Was Schirra airbourne or not? If the Titan had risen from the pad even an inch or two, the massive booster was even now sinking back to earth, where it would crumble, fall to the side, and explode--not necessarily in that order. If it had not moved at all, it was still bolted firmly in place. <br /><br />There was a liftime of difference between the two situations. If the Titan had moved, Schirra must make an almost instantaneous decision to eject, and yank the seat ejection ring between his legs. If there had been no such motion, he should do nothing.<br /><br />In training it had been emphasized that the clock would not start until a plug at the base of the gantry had been pulled loose by the departing Titan. Tick! Tick!There was a clock in front of Wally's nose. What did he do? Nothing.What did Stafford do? Nothing. What should they have done? By strict interpretation of the rules, Schirra should have pulled the ring and ejected both of them. However years of flying had taught Wally that sometimes you had better believe the seat of your pants, or more accurately, all the kinesthetic impulses feeding into your brain from your eyes, ears, tendons, joints, and muscles.<br /><br />This time Wally believed his body, and he was absolutely right. The tail plug had dropped loose prematurely and sent a false signal to the cockpit. The malfunction detection system had reacted b

 

[syndroma]

Calli, just some clarifications on Soyuz 18.<br /><br />Soyuz Abort System can be triggered by following events:<br />1. LV going off course by 7 degrees.<br />2. Pressure drop in one of the chambers.<br />3. Nozzle gimbaling mechanism set in extreme position.<br />4. Command from the ground.<br /><br />The second stage of the Soyuz 18 separated properly, but the tail fairing of the third stage failed to do so. Because of that the LV started uncontrolled roll along its axis. But Rule #1 was blocked for 10 seconds after the separation (a separation can cause brief deviation larger than 7 degrees). When that 10 seconds elapsed, Rule #1 triggered the abort. But the spacecraft already rolled upside down, while onboard computer thought it's just 7 degrees deviation. That caused the descent module to experience negative aerodynamical force and huge G-loads. The trajectory was so curved that it even landed much to the west of the second stage!

 

[CalliArcale]

Cool! I didn't know the details of that. <img src="/images/icons/wink.gif" /> So it wasn't that the second stage had failed to separate but that the transstage had failed to separate. <div class="Discussion_UserSignature"> <p> </p><p><font color="#666699"><em>"People assume that time is a strict progression of cause to effect, but actually from a non-linear, non-subjective viewpoint it's more like a big ball of wibbly wobbly . . . timey wimey . . . stuff."</em>  -- The Tenth Doctor, "Blink"</font></p> </div>

 

[jamie_young]

Thanks Calliarcade. Yes, I agree with your post. The point on where an experienced astronaut might foresee a problem before a computer is programed to know is something I hadn't thought of. At least it looks like the option is firmly part of the CEV plans and that is good. No more Challengers, although I know there was nothing that could be done on that sort of vehicle <img src="/images/icons/frown.gif" />