NASA's Kepler Enters Emergency Mode 75 Million Miles From Earth (theverge.com) 104
Loren Grush, writing for The Verge: NASA engineers have declared a mission emergency for the agency's planet-hunting spacecraft Kepler, which has somehow switched into emergency mode. Now that a mission emergency has been declared, the Kepler team has priority access to NASA's deep space telecommunications system in order to try to get the spacecraft back to normal operations. Emergency mode is the lowest operational mode the spacecraft has. It also requires a lot more fuel than usual, which is why the Kepler mission team is working hard to get the spacecraft back to normal. But communication with Kepler isn't easy. The spacecraft is estimated to be 75 million miles away from Earth right now, according to NASA, so any communications signal traveling at the speed of light will take up to 13 minutes to travel to and from the spacecraft. Kepler has detected nearly 5,000 exoplanets over the years -- of which 1,000 have been confirmed.
Re: (Score:1)
Pretty sure they wrote it all in Ruby on Rockets.
Re: (Score:1)
Yeah, but it uses systemd as well.
Re: (Score:2)
Then we should be all set. It will boot up even faster, now!
Re: (Score:3)
Typically spacecraft software is computer verified for its correctness -- the main reason it is so expensive. So using Rust would not have helped.
Re: (Score:2)
A "computer" can not "verify" software for correctness.
Aka, halting problem and such.
Re: (Score:3)
Correction: A "computer" can not "verify" *all* software for correctness.
It makes quite a big difference, there is an entire field devoted to Formal Verification.
Re: (Score:2)
Yes, I simplified.
Formal Verification requires a program and a specification (in a formal language like Z) and only proves that the program is conforming to the specification or not. In other words it does not prove that the program does what it is intended to do.
So: for twice the work, once formulating the concepts in Z and then coding them in a particular language you now can run a a formal verification, either manually or with computer assistance.
However: with double the work you have now double the opti
Re: (Score:2)
You are still over-simplifying somewhat.
The option to make a mistake program and in the specification are not exactly the same. Bugs in the code and bugs in the spec are two different kinds of mistake. One has a lower probability and is more visible than the other. Rather than doubling the number of places that a bug can be made, it remove the options for implementation errors and replaces then with fewer options to make specification errors.
Assuming that specs are being written manually. It is far more use
Re: (Score:2)
So now we are quite a long way from what you originally wrote?
Actually not.
As a automated verifier only can check if a program confirms to the specification.
With manual verifying like Hoars method etc. you can "figure" what the program is doing. And then make a rational choice if that makes sense. An algorithm can hardly do that.
Re: (Score:2)
Typically spacecraft software is hand verified for its correctness -- the main reason it is so expensive.
On a space program, the teams are highly experienced at verifying software.
Re: (Score:2)
Its code of conduct simply forbids bugs. Awesome.
Re: (Score:1)
Its code of conduct simply forbids bugs. Awesome.
When the NASA engineer stated that "failure is not an option", he was really talking to the computers.
Re: (Score:2)
Jokes aside, there was an interesting presentation by Mark Maimone at CppCon 14 about using C++ on the Mars rovers:
https://www.youtube.com/watch?... [youtube.com]
They use a fairly restricted set of C++ features, but overloaded new and delete operators to avoid fatal memory space issues.
Why more fuel than usual? (Score:1)
> Emergency mode is the lowest operational mode the spacecraft has. It also requires a lot more fuel than usual
Why?
Re:Why more fuel than usual? (Score:5, Informative)
Re: (Score:2, Insightful)
What's a reaction wheel then? Some kind of gyro system?
What kind of internet access do you have that lets you get to Slashdot, but not Wikipedia?
Hey, mod this comment down, too (Score:2, Insightful)
This is fucking Slashdot, not a youtube comments section. We should not be coddling people who can't figure out how to use the internets.
Re: (Score:2)
Who is "we"? Why import Stack Exchange jackassery? Does the mod system have utility words for self-appointed board monitors to bitch at innocent questions?
Re: (Score:1)
Who is "we"? Why import Stack Exchange jackassery?
If you are too lazy to check Wikipedia to see if your exact terms match the title of an article, then fuck you.
Re: (Score:1)
Re:Why more fuel than usual? (Score:5, Informative)
A gyro to rotate you, not sense your rotation - here are the Kepler [astrobites.org] reaction wheels. (It has already had 2 reaction wheel failures.)
Note: Inertial sensing gyros are generally gimbaled, so they can stay rotating in the same direction. Reaction wheels are generally fix mounted, and are spun up and down as needed to get the desired attitude (or rate of rotation). If the spacecraft is being torqued by something (say, a small gas leak), the wheels will spin faster and faster to soak up the excess angular momentum until they reach their design limit, and have to be despun. This is called a momentum dump, and requires some other system (i.e., thrusters) to finally get rid of the excess angular momentum.
Re: (Score:2)
Inertial sensing gyros, in the year 2016, *are not gimballed*. Almost every available unit for the last 30 years has been a strapdown system. where the gyros remain fixed to the body (and are controlled by a rebalance system to keep it that way), output a delta-angle, which is then integrated to determine the attitude.
Re: (Score:2)
So, there are a couple of spacecraft, if not more, still using gimballed gyros for inertial sensing?
Also ... integration has at least one potential cause of problems. (I saw this in inertial guidance for an ROV recently, which they use to count turns on the daughtercraft umbilical. Quite an important parameter.) If the original software design didn't envisage completing 1000 revolutions (pick a number) and that code is re-used 1
Re: (Score:2)
From the Wiki page, Kepler was initially planned to launch in 2006, so design and hardware freeze would have been 2003 if not earlier. It takes a long time to get hardware into space. IIRC, the last (as in final) servicing mission to the Hubble installed a 486-class processor to upgrade the 386-class processor that had been working for the previous 20-odd years.
Re: (Score:3)
Basically, it's a wheel that you can speed up or down with a motor. The axis of the wheel is fixed to the spacecraft body. When you accelerate the wheel rotation, its angular momentum increase. But as the total angular momentum of the spacecraft does not change, the rest must rotate in the other direction, by reaction. Typically you include several of them in one spacecraft because each of them only control one axis.
Re: (Score:2)
Particularly in this case, since two reaction wheels have previously failed.
Re: (Score:3)
From the SOHO spacecraft web page [nasa.gov]:
In other words, the spacecraft goes into a mode whe
"is currently 75M miles away right now"? (Score:4, Informative)
You know, if you say "is", you really don't have to say "currently" or "right now", much less both.
Re: (Score:1)
Wrong: "is" by itself can often imply a continuing situation. If the distance varies (e.g. not orbiting the earth), adding the extra verbiage implies that the distance varies. The problem with the way they wrote it is that they didn't explain why the distance would be variable.
Re: (Score:2)
The Kepler spacecraft is on a Sun orbit that falls behind the Earth a certain amount each year, so it is a variable number, not a set number.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Re: (Score:2)
but it is aliens.
Re: (Score:1)
"Depends on what your definition of 'is' is": Bill Clinton
Re: (Score:2)
Re: (Score:2)
Even if "currently" is not technically needed, it helps because this thing is moving away from Earth, if slowly, falling behind by 6 days or so per year.
Re: (Score:2)
Well, relatively speaking... Yes. ;-)
Re: (Score:3, Interesting)
Re: (Score:2)
Re: (Score:2)
Fun fact: right after the French Revolution, there really was a decimal time standard, with a 10-hour day, a 100-minute hour, and a 100-second minute.
Re: Math Fail (Score:4, Interesting)
To & from is 150 million miles, which works out to 13 minutes. How did they 'fail'?
Re: (Score:2)
To & from is 150 million miles, which works out to 13 minutes. How did they 'fail'?
It's actually an English failure, not one of mathematics. The statement is vague, and doesn't make it clear whether they're talking about a round trip, because it talks about "a signal" when in reality if you have a round trip, you have at least two signals.
Re: (Score:2)
I thought something seemed off as it takes 8 minutes for sunlight to reach us and the Sun is 93 Million Miles away.
Re: (Score:2)
Re: (Score:2)
true, but still one restart should do it. Not a whole series of back and forth single responses. I'd think the craft would be sending a stream of status information anyway assisting in pinpointing the issue.
Re: (Score:2)
Re: (Score:2)
true, but still one restart should do it. Not a whole series of back and forth single responses. I'd think the craft would be sending a stream of status information anyway assisting in pinpointing the issue.
No, because you have to figure out what went wrong before you try and issue such commands. The first rule is "do not lose the spacecraft," which can (and has) happened if the wrong commands are uploaded.
The spacecraft will send back a bunch of status info (bandwidth is limited, so that typically won't be everything), the spacecraft engineering team puzzles over that, and then send commands such as "try doing X, and then send us back the readings from sensors Y and Z." 15 minutes+ later they get the results
Re: (Score:2)
true, but still one restart should do it. Not a whole series of back and forth single responses. I'd think the craft would be sending a stream of status information anyway assisting in pinpointing the issue.
I doubt the computer system is the problem here, or that it crashed. The problem is more likely an unexpected response to a command for the spacecraft to do something.
To use a metaphor, this is not like a robot with a CPU crash, this is like a robot reporting that it stopped moving because can't tell what its leg is doing.
Re: (Score:1)
13 minutes? More like 7 unless there's a lot of processing delay. Batch Jobs perhaps? Sorting cards?
People like you who are so eager to show how clever you think you are, that you'll find fault where there is none instead of trying to understand how something might be possible, are what's wrong with this site. You've made it a hostile environment for anyone who actually wants to communicate, especially something non-trivial. Anyone doing so will spend more time explaining how you've twisted and misunderstood what they wrote than actually conversing.
That you've become so common is part of the reason why
Re: (Score:2)
Still not as bad as StackExchange dot assholes.
Re: Math Fail (Score:1)
Your comment just made me notice,
Why the hell don't sigs show up on the mobile site?????
Re: Math Fail (Score:5, Funny)
Why the hell don't sigs show up on the mobile site?????
Because, sigs is short for signature, and signatures were originally on paper...aka STATIONARY. Therefore, they can't be mobile.
Re: (Score:2)
Probably second best joke ever, best one still goes to phantomfive, I believe. :-/
To bad no one will get it
Re: (Score:2)
NASA are working on that. https://en.wikipedia.org/wiki/... [wikipedia.org]
Kepler has been really impressive (Score:5, Insightful)
Even if they don't get ti fixed, Kepler has had an absolutely amazing run. The initial planned mission lifetime was 3.5 years, and that was in 2009. So we've gotten almost twice as much out of it as it was planned.
One of my favorite computer games from the 1990s was Masters of Orion II, 4X space exploration/conquering game. One thing in that game and many similar games was the idea that you couldn't find out what planets were in a star system until you had actually sent a probe there. It is absolutely amazing that shortly after those games were made, we had the technology to detect planets in other star systems while remaining in comfort here.
Re: (Score:2)
That's a lie (or ignorance). It has been a mission plagued with problems that has not met expectations. I don't know if it has fulfilled it's planned mission, but if so, it was only just recently:
Re: (Score:2)
It is a matter of interpretation. He is neither a liar nor ignorant. Having detected 1000 and a potential 4000 more exceeds NASA's site's stated goal of hundreds of planets.
Re: (Score:2)
That's a lie (or ignorance). It has been a mission plagued with problems that has not met expectations.
Go straight in with 'lie' eh? Must be an interesting world you live in. Anyway - Kepler has confirmed over 1000 exoplanets and has four and a half thousand candidates. That's a successful mission. We didn't know that planets (particularly the small, non-hot-jupiter ones) were this prevalent until Kepler.
Re: (Score:2)
Re: (Score:2)
We can't visit them but we could send messages to promising ones.
Re: (Score:2)
who gives a shit about the Earth's moon and planets, we'll never be able to visit them.....
-- some moron like you who doesn't know what is technically possible
Re: (Score:2)
More importantly, it's the people who do give a shit, and who don't really care about what is "technically impossible," because that only encourages the invention of new technologies. Remember - even into the 1940s, people were decrying Goddard as an idiot for thinking you could reach orbit with a liquid-fuelled rocket.
Re: (Score:3)
if you want to call those that seek scientific information about the universe nutters go ahead; if you want to jerk off at the end of life of successful monumental project go ahead, you're just being a jerk-off troll.
kepler already did its job and from backlog of data which is still being processed we'll know percentages of what types of stars have what kinds of planets, distributions of rocky and gaseous planets, percent planets in habitable zone and also gas giants in habitable zone.
Re: (Score:2)
So what? You're not going there. Ever. Nobody is going anywhere.
Just because you haven't been able to leave your mom's basement, doesn't mean others are so restricted. The only thing slowing space exploration currently is political will, not technological progress. If the human race worked together to really make an effort at space, we could have orbiting colonies in around 20 years, it isn't that hard.
Re: (Score:2)
You're right, no one from Earth will ever walk on another world such as the moon. Puny minded little turd, aren't you?
Reaction wheel failure? (Score:2)
Kepler already had 2 of its 4 reaction wheels fail. If a third is gone, it'd mean they have to use the thrusters more, reducing mission lifetime.
Re:Reaction wheel failure? (Score:4, Informative)
Given that we know that Kepler has a bad batch of reaction wheels, that two of the four have already failed [astrobites.org], and that this emergency mode happened while the spacecraft was being repointed to the Galactic Center for a microlensing campaign [nasa.gov], which inevitably would mean a lot of reaction wheel use, I very much fear that this means that another reaction wheel has failed and the K2 mission is over.
Some aliens ... (Score:5, Funny)
Re: (Score:1)
That makes it sound like a Borg put tape on it to watch it struggle like a cat with tape on its paws.
detected the wrong planet (Score:2)
Must have detected a planet that didn't want to be detected.
UPDATE -- EM over... (Score:3, Informative)
Better now (Score:1)
Kepler is now out of Emergency Mode and responding to commands.
The Event happened before the manuver was started and probably did not involve the reaction wheels.
The actual cause is not known yet.