WASHINGTON – On Aug. 5, 2012, after completing a nine-month, 354 million mile journey, mission controllers at NASA’s Jet Propulsion Laboratory will attempt to break the fall of a rover-carrying spacecraft traveling at about 13,000 miles per hour to zero and set down the rover in a Martian crater.
The landing sequence for the Mars rover Curiosity is called the “Seven Minutes of Terror.”
For the first time, NASA will use an untried descent and landing technique to lower Curiosity to the surface near Gale Crater using a sky crane. The entire landing sequence will be controlled by computers relying about 500,000 lines of code that will control no less than 76 rockets and thrusters.Mission managers won’t know whether it worked for 14 minutes, the time it takes a signal from Mars to travel back to Earth.
To the casual observer, using a sky crane to land on Mars may look crazy, but “it is the result of reasoned engineering thought,” argues JPL’s Adam Steltzner, chief engineer for the Mars Science Laboratory.
“If any one thing doesn’t work just right, it’s game over,” adds JPL’s Tom Rivellini.
Previous Martian landings used what amounted to air bags to break the fall of rovers. Curiosity is too big and heavy – about the size of an SUV – to use that landing method. Hence, JPL engineers came up with a system of braking rockets, a huge parachute and the sky crane to slow Curiosity’s fiery entry into the faint Martian atmosphere and gently deposit it in Gale Crater, which is suspected of harboring signs of Martian life.
The JPL engineers and scientists are sticking their necks way out on this mission, and that’s precisely the type of technological risk-taking that is needed to explore the solar system with our marvelous machines.
Could the 14-minute waiting time be compared to the anticipation that the Cape Canaveral control room experienced waiting for the Apollo missions to the moon back in the 1960s? Seems then there was also a waiting period before confirmation that the Eagle had landed. Maybe this is much more hazardous, and good that it is unmanned.http://www.youtube.com/watch?v=6R3j1NU2nQM
Comms delays for Apollo was more like 2 seconds, and there was constant telemetry flowing back to Houston mission control during lunar descent. Indeed, the LM crews were always being told to realign their antennas so Houston could continue receiving systems telemetry and voice. The difference of course was that there were two humans aboard. We're a long way from a manned Martian landing.
I was at Aerojet in Sacramento during the moon landings. Aerojet produced the engine (that fired behind the moon) to bring the men back from moon orbit. The live feed was put on the Aerojet phone system, and we all sat there holding our breaths. There was a voice-over countdown that reached zero a couple of seconds before we heard the signal from the capsule. Perhaps those were the 2-second telemetry delay (while we sat there without oxygen to our brains for 2 seconds). The engine never failed, but perhaps some of our hearts did.
That is a lot of unproven technologies to try all on the same mission. The reasoning is logical, but the end result could be a catastrophic waste of time and money. I'd be interested to hear what JPL's engineers calculated was the probability of success.
Personally, I would want samples from the crater. Data of Mars would be advanced quickly, from the back splash of magma displacement. Look for unique elements from the planet and the striking body to correlate with future samples.
And that's before you leave the crater.
I understand the desire to get larger rovers to the surface, but if this requires controlled descent, rather than ballistic descent, then wouldn't it be prudent to first achieve some mundane infrastructure missions? Instrumentation and comms satellites have been sent, but why not a Mars global positioning system? I would think that would greatly simplify navigation, both for the entry vehicle and for the rover, as well as provide precise location of any interesting discoveries. Its cost would be amortized across many missions for decades.
There are metric units, and there are obsolete units.
I was slightly perturbed to hear the parachute engineer still referring to "pounds-force".
I'm English, and we don't use "English" units, so why does anyone else?
Is it irony or subtle humor that the graphic at about 0:21 says "500,000 LINES OF [sic] COD3", followed by "ZERO MARGIN OF ERROR"? Just hoping we're not back to the era of the km/miles debacle and sensors on spring-loaded legs without debouncing.
"...then gently deposit it on it's wheels on the surface." (ie. deploy at 20m then descend to 7m)
Looks easy in the animation... (haha) Liked the way he added "...on its wheels..." Wonder how they are going to determine when it is on the ground and not high centered on a big boulder...
The complexity of this undertaking is mind boggling. Lowering the rover down on a tether from a rocket propelled remote controlled vehicle? Indeed, no evidence of the K.I.S.S. principle here. That 14 minute delay to find out if all went according to plan is going to be brutal at mission control.
Great video, BTW.
Looking at what you've mentioned about landing the rover using a rocket propelled "remote controlled" sky crane, a question that occur to me:
Is it the manual control from Earth to control the sky crane? If yes, then how much delay in the response of that "sky crane" will be after the commands are sent? This looks not possible to me.
It is not remote controlled. The entire sequence is completely under autonomous computer control. This is mentioned in the video. All high speed manuevers at such distances must be autonomous because the round trip comms delay is nearly half an hour.
It is interesting to know exactly at what instance the landing sequence is activated and will it be activated from the control center on earth? the 14 minute delay in the signal transmission makes the whole maneuver very interesting.
There is not a lot of choice. From what I understand, it is not going into Mars orbit to await a command, but rather a series of 5 trajectory correction burns will aim it at the atmospheric insertion point to within 2 km at a (Mars relative) speed of about 20,000 km/h.
Another first is that it will not traverse the atmosphere ballistically like previous landers, but will use thrusters and its cone shaped shell to enable attitude adjustments and to perform a series of s-shaped turns to adjust its flight path (and perhaps to bleed off speed?) There are an uncomfortable number of "firsts" in this mission.
It all automation and robotic work at high velocity and heat, it would be very much interesting event to see the landing process, but unfortunately no camera can record the actual landing process and send it back to earth. But still it is very good that people who are working on this project has disclosed this much details of the project. The comment trails are also very much interesting, I would like to see more interesting comments coming up here.
Personally, I applaud the engineers behind this entire mission. They were given an almost impossible task and came up with a solution (despite all the arm chair quarter backing going on). Though there are so many variables involved anything could happen (one slight crack in a PCB trace is all it takes to destroy an entire mission) they undertook it all with eyes wide open and didn't appear to back down in the slightest (as an aside, I once was given the task to design a power transistor for a flight surface of a fighter jet. I begged off the project). And let's be careful of something I see creeping more and more into into the "common consciousness"; there are "engineers" and then there are "Engineers". These people are the best of the best (I won't get dragged into silly arguments about that statement). This isn't a new iPhone we're talking about (antenna where?); these guys and women have the brains and the tools to do the job. A friend from the Air Force made me laugh-out-loud once when he was describing fighter pilots; "they're smarter than you, they think faster than you, they're taller than you, they're teeth are whiter and straighter than yours, they are simply better made humans than you". Let's not assume that the people working on this are the same as the guy in the next cubicle that thinks that current flow is "left to right". I hope they pull pull this off; if the do there should be ticker tape parades and a holiday for them. Instead, we'll most likely have to hear more about the Kardashians.
Well said, mtripoli. The folks at JPL were given a problem and came up with the best possible solution. It very well may not work, but we will learn a lot in attempting it. If the recent SpaceX mission to the space station is any indication, software verification must have been exacting. As noted earlier, there are plenty of unknowns during the descent. If we fail, we fail with the whole world watching. That's the strength of our space program.
A word about automation and robotics: I was at a talk regarding the "state-of-the-state" in automobile (let's say) "smarts". After a demonstration (I don't want to be coy about it, I can't go into details as to what I saw) the statement was made "the biggest problem with automobiles and traffic is the person behind the wheel". Of course this statement implied major changes to our roadways and highways but the point was well taken. Things are a little different than they were during the Apollo missions (http://en.wikipedia.org/wiki/Apollo_Guidance_Computer).
If we take a look at Space-X and at this mission, we can see what I see as the future of space leadership for this country: Private industry refining orbital work into just another commercial business and NASA exploring, inventing and pushing the envelope.
Space technology works because it is an assembly of simple components or other simple assemblies. Each assembly step is examined to ensure it does not create additional complexity in its components or other assemblies (loose coupling, encapsulation). Components are based on known provable or reliably simulated principles (= reusing prior art). Interfaces are designed and reused consistently, so that the parent design can be exhaustively tested and the benefits inherited by the child applications.
Can we learn from this? It _is_ rocket science, but it is mostly not complicated in the sense of magic, impenetrable black boxes.
On the other hand, a lot of the math IS complicated, in the flight vectors and even the reliability engineering statistics.
This is the most complicated automated high-reliability project I've ever heard of. I think it's great, and I wish them complete success.
This is why we shouldn't waste money on manned missions. In order to get a big science package to Mars, they're doing something incredibly risky.
If humans were sent, the whole project would be about keeping them alive. The project would take forever, and in the end, there'd be no room left over for actual science.
My guess is that the first manned mission to Mars will be a one-way trip. We simply lack the propulsion technology to break out of the Martian "gravity well" and return home. This is controversial, but probably the only way humans will get to the Red Planet until we figure out a better form of propulsion that doesn't require dragging along millions of tons of fuel for the return flight.
One simple question was never addressed in this extraordinary story and challenge that is certainly attracting a record breaking volume of commentary. Why not build upon the manned lunar landing technology with a radar equipped control rather than a human pilot? Seems like a much lower risk solution for a comparable payload weight (the lunar lander weighed 16 tons but the moon has a lower gravitational force).
*Not singling out DrQuine - this is a generic statement*
One thing I've learned over the years; when it comes to large design and engineering programs like this, I assume anything I may have "thought" of, so did they, and got there much faster than I would have. As with any endeavor, or simply our day to day jobs, when you have your nose in it 24/7 you see things others around you couldn't if they wanted to. I'm all for design reviews; they are a vital part of the engineering practice. But doesn't it burn your behind when someone asks "Did you think of xxx"? In your head your saying "Yes, and for xxx reasons I abandoned that idea. If you were as focused on this as I am you would not have asked that question". This is not to say I work in a bubble, or don't value others thoughts and ideas. Just the opposite; I think I'm pretty good at what I do, evaluate all reasonable options before coming to a decision. If I get "stuck" I'll ask for input. Certainly there is no way JPL could have explained every aspect of every decision they made in a short video (or even a video 30 hours long). Given the resources available to them (I'd love to have 1% of the computing power available to them for FEA) I have to imagine they've examined every past mission that we know about (and those we don't) to come to the decisions they did.
I think the simple answer is that Mars is much farther from Earth than the moon. The Apollo landings had the advantage of Houston being able to monitor telemetry, critical LM systems and fuel consumption in near-real time. Due to the distance and comms delays, a Mars landing must be autonomous, and hopefully the system is designed to interpret onboard radar and adjust the flight profile "on the fly."
Very interesting article and discussion...but I am unable to enjoy the video. For whatever reason, not all of the controls are visible. I wanted to watch full-screen, so I clicked around, but no go. Now I get no sound.
I am using Firefox 13.0.1 up-to-date. I also tried IE9, but results are the same (no full-screen button visible, no sound).
Can this be fixed? Or can someone provide a link to somewhere else I can see the video directly?
We thought the video was was very relevant to our engineering audience because the JPL engineers were sticking their necks out and trying to come up with a solution. It's highly likely this won't work, but these guys are being honest and their mistakes will be public for all the world to see. That's the way we've always done it, and no one can touch us in terms of space exploration.
Very nice video. At least I can appreciate why MSL went way over budget. The first question that came to my mind - why did Curiosity have to be so big and heavy as to require such an elaborate landing method? When I think of the best value for our space dollar, I think of Spirit and Opportunity, which lasted years beyond their required mission life.
A reader asked earlier whether NASA had calculated the chances for successfully landing Curiosity on Mars in August. Here is the full response to our question from a spokesman at the Jet Propulsion Laboratory, which is subject to different interpretations:
"No. Important risks are the unknowns, which resist calculating."
From July 16 NASA press conference:
"We expect to get Curiosity safely onto the ground, but there is no guarantee. The risks are real," said Pete Theisinger, the Jet Propulsion Laboratory's Mars Science Laboratory project manager.
As we know, it landed successfully! CONGRATULATIONS TO EVERYONE AT JPL! I said it before and I'll say it again, you guys should have a holiday named after you and a ticket-tape parade! I can't begin to express how happy I am for you all! Congrats! Congrats! Congrats!
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.