SpaceX Falcon 9 CRS/SpX-10 Dragon - Feb. 19, 2017 - Discussion

[Jim]

Right. If the Process

What process?  That is the issue.

[ATPTourFan]

The Process that today recognizes that computer code and sensors aren't perfect and allows for a human-initiated abort window.

I guess I don't understand what the alternative is you would suggest they use.

[Comga]

The point is about processes

Correct and that is why Challenger and Columbia are no longer relevant comparisons.

In this case your expert opinion is still just opinion.
"Lessons Learned" are never learned by all, and never all learned. Those remain valid lessons, as do your points that some of the perceived differences aren't so different. It could be said that Challenger and Columbia are irrelevant because it's a warm day and there is no spray-on foam, bit we know that would be silly. 

[Jim]

The Process that today recognizes that computer code and sensors aren't perfect and allows for a human-initiated abort window.

I guess I don't understand what the alternative is you would suggest they use.

No, if this is an issue, then there should be hard and fast rules dealing with it and not gut feeling.  If it is not good to launch today, then it shouldn't be good to launch any other day.

[ATPTourFan]

The rules today indicated green to launch. But sometimes the real-world combination of factors is more than the code can handle, and it shows Green.

Understanding where weaknesses could potentially reside in your launch process, even if realized at T-15, still warrant an abort.

Should Elon stand by and say nothing if he feels the process has encountered a weakness given the unique scenario.

Processes are designed by flawed humans, very smart flawed humans. Over decades they get more robust, but can you ever just sit back and tell yourself -- it's green so we're good? Haven't there been failures (by everyone who ever endeavored to launch a rocket) in the past using that exact position?

[Norm Hartnett]

Today was as good a wet rehearsal as you can get. I'm sure that they learned a number of valuable lessons about process and the new pad today. I fail to see why the Chief Technical Officer shouldn't call abort if he sees the need to.

[Jim]

The rules today indicated green to launch.

The rules were flawed then, if they stopped the launch without following them.
Also, launch rules are not just computer code.  There are many that required human action and intervention.

[ATPTourFan]

The rules today indicated green to launch.

The rules were flawed, if they stopped the launch

So from what I'm hearing you say, SpaceX should have by now designed a "perfect" process/rules. How many other organizations thought they had a perfect process, followed rules, and still experienced failures?

Realizing that your process can have weak areas and taking action, and permanent corrective action, is in my opinion smart. There is no perfect set of rules, no perfect process. If there was we'd have one button and could sit back and watch the magic without scrutinizing dozens of computer screens constantly throughout the mission.

[southshore26]

The rules today indicated green to launch.

The rules were flawed, if they stopped the launch

So from what I'm hearing you say, SpaceX should have by now designed a "perfect" process/rules. How many other organizations thought they had a perfect process, followed rules, and still experienced failures?

Realizing that your process can have weak areas and taking action, and permanent corrective action, is in my opinion smart. There is no perfect set of rules, no perfect process. If there was we'd have one button and could sit back and watch the magic without scrutinizing dozens of computer screens constantly throughout the mission.

Let it go... you're not going to change his mind... even if he's wrong he's right.

[Jim]

So from what I'm hearing you say, SpaceX should have by now designed a "perfect" process/rules. How many other organizations thought they had a perfect process, followed rules, and still experienced failures?

Realizing that your process can have weak areas and taking action, and permanent corrective action, is in my opinion smart. There is no perfect set of rules, no perfect process. If there was we'd have one button and could sit back and watch the magic without scrutinizing dozens of computer screens constantly throughout the mission.

Again, what changed from the time they discovered the issue until T-13 seconds?  If it wasn't good at T-13, why was it good earlier and they continued the count?

[rayleighscatter]

The rules today indicated green to launch.

The rules were flawed, if they stopped the launch

So from what I'm hearing you say, SpaceX should have by now designed a "perfect" process/rules. How many other organizations thought they had a perfect process, followed rules, and still experienced failures?

Realizing that your process can have weak areas and taking action, and permanent corrective action, is in my opinion smart. There is no perfect set of rules, no perfect process. If there was we'd have one button and could sit back and watch the magic without scrutinizing dozens of computer screens constantly throughout the mission.

It's obviously not a perfect set of rules as they apparently don't even indicate any sort of clear chain of command in mission control. Or at least that the flight director is either untrusted or a toothless position.

[ATPTourFan]

Again, what changed from the time they discovered the issue until T-13 seconds?  If it wasn't good at T-13, why was it good earlier and they continued the count?

I thought you told us. It was "good" earlier because the process indicated "green". However, smart people were likely throughout the count looking deeper into potential fault trees and relationship between the issues they had experienced with this vehicle and determined they'd prefer to look at it rather than bet their process was perfect.

Elon (and likely his senior group) hadn't satisfactorily convinced themselves that this issue was worth risking the mission, even if the computer said it was good enough.

Because the process and people who designed it aren't perfect, you build into the process a means for someone to say "HOLD HOLD HOLD" because they don't feel something is right.

Your options are:
1) build a perfect process and install the big GO button
2) recognize your process will never, and can never be perfect and build into it the ability to hold/abort

Action taken today to HOLD (and thus abort the instantaneous window) wasn't taken lightly. Do you think it was?

Also, just because Elon tweeted he "called the hold" doesn't mean that that Flight didn't concur and stop the clock. We don't know that. You are assuming no chain of command.

[Johnnyhinbos]

The rules today indicated green to launch.

The rules were flawed then, if they stopped the launch without following them.
Also, launch rules are not just computer code.  There are many that required human action and intervention.

I could be wrong, but it was my understanding that the TVC was a known issue, but one that could have had at least some probability of being corrected before T-0. Therefore the launch countdown was allowed to proceed as normal. When it was determined at T-13 seconds that the TVC data was still out of the acceptable window the launch was scrubbed.

This doesn't sound like a flawed protocol, rather this sounds like a prudent path. I'm not sure as to why this is even a point of contention?

[ATPTourFan]

Johnny, I've just tapped you into the ring. I'm out.

Goodnight everyone. Let's hope for a successful, safe launch tomorrow.

[rsdavis9]

If I were them(elon and people). It would come down to something that is okay for the computer but seems a little off and I don't understand completely why it is a little off then I would call it off until I do understand it. I am a programmer and I don't feel comfortable until I understand the reason my code fix fixed the problem.

[Lar]

lets not do Jim vs the world.

If someone experienced calls a manual hold that is an opportunity for process improvement. Leave it at that.

[llanitedave]

The rules today indicated green to launch.

The rules were flawed then, if they stopped the launch without following them.
Also, launch rules are not just computer code.  There are many that required human action and intervention.

Of course the rules were flawed.  That's why Elon stopped the launch.  If the rules covered all relevant contingencies, then the board would not have been green.  Calling a manual halt to this is an acknowledgement that there are still some details to work out in the process, and that's exactly what they'll spend this time doing.

All rules, all processes are flawed in some way.  There's no shame in admitting that and acting on that fact when things don't look right otherwise.  I don' know why Jim is making this into an attack.

[rayleighscatter]

Also, just because Elon tweeted he "called the hold" doesn't mean that that Flight didn't concur and stop the clock. We don't know that. You are assuming no chain of command.

Then flight director should have "called the hold."

Musk doesn't take part in each flight, and he doesn't take part in each training evolution with mission control so when he suddenly steps in and usurps control it sends some real mixed messages. When a split second decision matters now who does the person at the console turn to?

The decision making process has to be clear and consistent for safe operations. This may not have been a critical operation but what lesson does the flight team take in the future when a crew capsule is reentering and a split second decision has to be made and you hear the CEO in one ear telling you one thing and the director in your other ear telling you another?

[Rocket Science]

I have insufficient data to have an opinion one way or another, so in the absence of facts I'll just hold fire at this point...

[Lee Jay]

So from what I'm hearing you say, SpaceX should have by now designed a "perfect" process/rules. How many other organizations thought they had a perfect process, followed rules, and still experienced failures?

Realizing that your process can have weak areas and taking action, and permanent corrective action, is in my opinion smart. There is no perfect set of rules, no perfect process. If there was we'd have one button and could sit back and watch the magic without scrutinizing dozens of computer screens constantly throughout the mission.

Again, what changed from the time they discovered the issue until T-13 seconds?  If it wasn't good at T-13, why was it good earlier and they continued the count?

When ULA finds an anomaly, and goes to the anomaly net to discuss it, isn't that because there isn't a "hard and fast" rule that a scrub must be called?  I'm remembering them adding a timer, or something, and continuing the count, and I'm remembering them saying something like "they got comfortable" with the anomaly and proceeded to launch.

I don't know what goes on when they get onto the anomaly net, but it sure sounds like there are humans making decisions rather than just following a flight rule and either scrubbing or not.

No?