Three spacecraft heading to Mars this summer(nytimes.com)
nytimes.com
Three spacecraft heading to Mars this summer
https://www.nytimes.com/interactive/2020/science/mars-perseverance-tianwen-hope.html
44 comments
I am wondering why NASA only sends one rover. Considering the investments into development, planning, and control and the relatively low launch costs nowadays, I would think it logical to send two or more probes to mitigate the risk of catastrophic mistakes.
"First rule in government spending: why build one when you can have two at twice the price? Only, this one can be kept secret. Controlled by Americans, built by the Japanese subcontractors. Who, also, happen to be, recently acquired, wholly-owned subsidiaries..."
https://www.imdb.com/title/tt0118884/quotes/qt1542424
https://www.imdb.com/title/tt0118884/quotes/qt1542424
A major point missed by the other answers is that these rovers are exceptionally specialized for their mission, so building a second rover wouldn't be able to accomplish many additional scientific goals, without major modifications. The Curiosity rover and Perseverance share a lot of components, but pretty much the entire science suite is replaced or upgraded. As such, the expensive parts weren't shared, even though the huge value of the bus/bogie system and the skycrane were shared. Even the EDL systems have been changed a lot; the entry system on Perseverance does localization during some phases with a camera (a first for a spacecraft afaic, maybe the shuttle counts though since a human eyeball steered it to a runway). If you don't have additional science objectives that a second rover could accomplish, it will be very hard to justify that budget, even if it is only 20% more (for example).
This being said, clever mission design might take advantage of this cost structure where multiple units are cheaper to build, intending to build two cheaper spacecraft instead of one expensive one; eggs in two baskets. In this case, we've seen this happen with spirit and opportunity, which was a wildly successful program.
This being said, clever mission design might take advantage of this cost structure where multiple units are cheaper to build, intending to build two cheaper spacecraft instead of one expensive one; eggs in two baskets. In this case, we've seen this happen with spirit and opportunity, which was a wildly successful program.
Same science suite in different landing areas would yield different results wouldn't it?
At what cost and how different? would the benefit from different results offset the cost?
NASA is pretty budget constrained. But they also admit to a risk-adverse culture. Part and parcel of having a safety first mission oriented to human spaceflight. I think your right and they would agree. We'll probably take much more risk very shortly, or fund entrepreneurial partnerships willing to do so. SpaceX has changed the game.
The other thing is if you look at the choice of Jezero Crater as the landing zone for the Mars 2020 Perseverance mission. The quality of high resolution astrogeological maps, atmospheric surveys, and even computer simulations is good enough today that we can predict right here on earth where evidence of fossilized life may be found, where rich deposits of silica aerogel may be located for terraforming, where the mineral deposits of niobium and tantalum are likely to be.
The other thing is if you look at the choice of Jezero Crater as the landing zone for the Mars 2020 Perseverance mission. The quality of high resolution astrogeological maps, atmospheric surveys, and even computer simulations is good enough today that we can predict right here on earth where evidence of fossilized life may be found, where rich deposits of silica aerogel may be located for terraforming, where the mineral deposits of niobium and tantalum are likely to be.
The launch cost isn't low. They're using an Atlas V-541 from UAL. The program cost is equivalent to about 10% of NASA's annual budget, which is a lot. Mars and the rovers are just one corner of NASA's focus.
It would cost a minimum of another $2 billion to duplicate this rover and mission, if you can find enough qualified people to task to it so as to not harm (by dilution) your ability to deliver on at least one of them. NASA's resources and talent pool are both very much limited.
It would cost a minimum of another $2 billion to duplicate this rover and mission, if you can find enough qualified people to task to it so as to not harm (by dilution) your ability to deliver on at least one of them. NASA's resources and talent pool are both very much limited.
This is exactly what they did with Opportunity and Spirit: https://en.m.wikipedia.org/wiki/Mars_Exploration_Rover
I think is is largely due to budget reasons, though.
I think is is largely due to budget reasons, though.
Those were solar powered. RTGs are not as plentiful as solar panels. Sending an RTG on rover B means some other space mission doesn't happen because they cannot get an RTG.
That's a good point. I forgot about the RTG shortage.
Sometimes NASA does send duplicates, including for Rovers. Spirit and Opportunity were identical. If all missions had a duplicate, we would have spent money to build an additional 15 spacecraft, and that would have only saved zero to two missions.
I'd guess it boils down to the relative expected contributions of random errors/failures vs. systematic errors/failures. Many failures were design flaws that would have created identical failures in any duplicate spacecraft of the same design. In that case, spending the money on two separate missions with unique hardware would have a higher return on investment than spending twice the money on one mission where either both spacecraft succeed and collect somewhat redundant data, or both fail and collect nothing.
Looks like NASA has attempted 28 missions to Mars, and four of these involved Rovers. All 4 Rovers have been successful. If NASA sent two rovers together each time, they'd simply have spent 1.5x money and completed the same # of missions.
Out of all 28 Mars missions NASA attempted, only 6 had some kind of failure:
- Mariner 3 and Mariner 4 were identical probes launched 23 days apart. Mariner 3 failed. We never figured out the root cause of Mariner 3 failure. Mariner 4 worked just fine.
- Mariner 8 and Mariner 9 were identical orbiters launched 21 days apart. Mariner 8 failed due to a botched installation of a diode onto a PCB which damaged that diode. Mariner 9 was successful.
- Mars observer failed. Root cause was never 100% definitively determined but NASA's final report focused on the hypothesis "that 2 grams of nitrogen tetroxide slowly leaked through a valve during the 11 months the Mars Observer was in space. When the fuel system was pressurised, the nitrogen tetroxide mixed with monomethyl hydrazine – an explosive combination....If NASA’s managers had studied the design well enough to understand its limitations, they could have taken steps to prevent the failure by periodically purging the fuel line." It sounds like a duplicate would have had the same problem -- increased costs for no benefit.
- Mars Climate Orbiter failed due to a software bug. Any duplicate mission definitely would have had the same bug and failed in exactly the same way -- increased costs for no benefit.
- Mars Polar Lander failed due to a software bug. It's not 100% certain that the bug would have been triggered by a duplicate mission, but it's reasonably likely. "The Failure Review Board concluded that the most likely cause of the mishap was a software error that incorrectly identified vibrations, caused by the deployment of the stowed legs, as surface touchdown. The resulting action by the spacecraft was the shutdown of the descent engines, while still likely 40 meters above the surface. Although it was known that leg deployment could create the false indication, the software's design instructions did not account for that eventuality."[1] I'd personally chalk a duplicate up to increased costs for no benefit, but an expert's interpretation may vary.
- Deep Space 2 failed, no root cause identified. 2 possibilities identified were design flaws, which could have affected a duplicate as well. 4 possibilities involved bad luck in landing so it's possible a duplicate would have had better luck.
So, if all missions had a duplicate, we would have spent money to build an additional 15 spacecraft, and that would have only saved zero to two missions.
[0] https://www.newscientist.com/article/mg14119080-300-cheap-de...
[1] https://en.wikipedia.org/wiki/Mars_Polar_Lander
I'd guess it boils down to the relative expected contributions of random errors/failures vs. systematic errors/failures. Many failures were design flaws that would have created identical failures in any duplicate spacecraft of the same design. In that case, spending the money on two separate missions with unique hardware would have a higher return on investment than spending twice the money on one mission where either both spacecraft succeed and collect somewhat redundant data, or both fail and collect nothing.
Looks like NASA has attempted 28 missions to Mars, and four of these involved Rovers. All 4 Rovers have been successful. If NASA sent two rovers together each time, they'd simply have spent 1.5x money and completed the same # of missions.
Out of all 28 Mars missions NASA attempted, only 6 had some kind of failure:
- Mariner 3 and Mariner 4 were identical probes launched 23 days apart. Mariner 3 failed. We never figured out the root cause of Mariner 3 failure. Mariner 4 worked just fine.
- Mariner 8 and Mariner 9 were identical orbiters launched 21 days apart. Mariner 8 failed due to a botched installation of a diode onto a PCB which damaged that diode. Mariner 9 was successful.
- Mars observer failed. Root cause was never 100% definitively determined but NASA's final report focused on the hypothesis "that 2 grams of nitrogen tetroxide slowly leaked through a valve during the 11 months the Mars Observer was in space. When the fuel system was pressurised, the nitrogen tetroxide mixed with monomethyl hydrazine – an explosive combination....If NASA’s managers had studied the design well enough to understand its limitations, they could have taken steps to prevent the failure by periodically purging the fuel line." It sounds like a duplicate would have had the same problem -- increased costs for no benefit.
- Mars Climate Orbiter failed due to a software bug. Any duplicate mission definitely would have had the same bug and failed in exactly the same way -- increased costs for no benefit.
- Mars Polar Lander failed due to a software bug. It's not 100% certain that the bug would have been triggered by a duplicate mission, but it's reasonably likely. "The Failure Review Board concluded that the most likely cause of the mishap was a software error that incorrectly identified vibrations, caused by the deployment of the stowed legs, as surface touchdown. The resulting action by the spacecraft was the shutdown of the descent engines, while still likely 40 meters above the surface. Although it was known that leg deployment could create the false indication, the software's design instructions did not account for that eventuality."[1] I'd personally chalk a duplicate up to increased costs for no benefit, but an expert's interpretation may vary.
- Deep Space 2 failed, no root cause identified. 2 possibilities identified were design flaws, which could have affected a duplicate as well. 4 possibilities involved bad luck in landing so it's possible a duplicate would have had better luck.
So, if all missions had a duplicate, we would have spent money to build an additional 15 spacecraft, and that would have only saved zero to two missions.
[0] https://www.newscientist.com/article/mg14119080-300-cheap-de...
[1] https://en.wikipedia.org/wiki/Mars_Polar_Lander
> Many failures were design flaws that would have created identical failures in any duplicate spacecraft of the same design.
I'm not so sure about that. Design flaws in aircraft are often not discovered until years pass, and it only affects one airplane (though the rest then get corrected). For a design flaw to cause a failure also requires the right conditions, which may be unlikely.
I'm not so sure about that. Design flaws in aircraft are often not discovered until years pass, and it only affects one airplane (though the rest then get corrected). For a design flaw to cause a failure also requires the right conditions, which may be unlikely.
Spacecraft are a little different because so many components can't be tested properly on the ground. "The pump stops pumping properly after 1 week with no gravity" is very hard to predict or test for.
My understanding is that there is a dedicated team for each probe on the ground, so two rovers = two teams, twice the cost to support. Or am I wrong?
The duplicate is almost always called EDU and stays in a cleanroom on the ground for troubleshooting.
Yes. Consider the Hubble space telescope. I would expect that building a pair rather than one wouldn't cost that much more than one, especially if they were built side by side.
This would likely have been far, far cheaper than the repair mission sent to fix the sole Hubble.
This would likely have been far, far cheaper than the repair mission sent to fix the sole Hubble.
Or two repair missions to fix the same defect in both telescopes would have cost even more
Budget justification. How government works.
> This would likely have been far, far cheaper than the repair mission sent to fix the sole Hubble
If two Hubbles had been built the same, both could have had the same implementation error [0] that required the first repair flight.
[Edit] Although I guess the second one could have been fixed on the ground based on experience from the first, assuming a non-simultaneous launch.
[0] https://en.wikipedia.org/wiki/Hubble_Space_Telescope#Flawed_...
If two Hubbles had been built the same, both could have had the same implementation error [0] that required the first repair flight.
[Edit] Although I guess the second one could have been fixed on the ground based on experience from the first, assuming a non-simultaneous launch.
[0] https://en.wikipedia.org/wiki/Hubble_Space_Telescope#Flawed_...
That's right. It would make sense to space the launches out a bit to be able to correct defects on the second.
Not the whole telescope, but they did build two mirrors, the backup one having been finished flawlessly by Kodak earlier than the one they used:
https://www.nytimes.com/1990/07/18/us/hubble-has-backup-mirr...
It ended up sitting “in its shipping crate in Danbury, Connecticut”: https://duckduckgo.com/?t=ffab&q=Danbury%2C+Connecticut&ia=m...
It ended up sitting “in its shipping crate in Danbury, Connecticut”: https://duckduckgo.com/?t=ffab&q=Danbury%2C+Connecticut&ia=m...
Well they did but it was classified for a while. Then the NRO donated it to NASA
https://en.m.wikipedia.org/wiki/2012_National_Reconnaissance...
https://en.m.wikipedia.org/wiki/2012_National_Reconnaissance...
I'm pretty sure what makes that stuff so expensive is all the Quality Control for every single part.
Its probably quite a linear cost vector they have.
Its probably quite a linear cost vector they have.
The failure to a "repaired PCB" stands out just scrap the damaged one would make more sense.
Being that this was 1971, it was probably more straightforward to repair the botched installation/repair rather than do the whole PCB over from scratch. Features on PCB's were much simpler then, not to mention much larger.
Their procedure was to test each component before installation only, in order to prevent any potential damage occurring during post-installation testing. So they evidently couldn't know to scrap it and replace the damaged one. After the Mariner 8 failure, they changed this testing policy to test PCB's/components again after components are installed on the PCB.
"A diode intended to protect the system from transient voltages was thought to have been damaged during repairs/installation of the pitch amplifier's printed circuit board, something that would not have been detected through bench tests." - Wikipedia, but I'm having difficulty finding any original sources that validate or expand upon this.
P.S. I think you may have responded to the wrong parent comment
Edit: "N72-13963" was a report given to Congress in October 1971 on the failure of the Mariner 8 mission, I can't find that report but it may have more details if you're curious enough to track it down.
Edit: I found a transcript of the report given to Congress, evidently in June of 1971: https://books.google.com/books?id=7FQVAAAAIAAJ
Dr. Low: The diode was external to this integrated circuit, mounted next to it. I don't have one with me, but the diode is another component about the same size. The integrated circuit which failed is known to be sensitive to electrical transients if it is not properly protected. And electrical transients large enough to damage an unprotected integrated circuit — caused, for example, by engine start, stage separation, et. cetera — are well within the specifications of the Atlas/Centaur vehicle and are clearly possible.
Because of this, the integrated circuit was protected with Zener diodes which in effect were designed to act as "safety valves" to prevent potentially damaging electrical transients from reaching the integrated circuit itself. In accordance with standard practice in the Atlas/ Centaur vehicle — as well as in many others — these Zener diodes were checked individually prior to their installations, but were not tested thereafter. This practice of not testing protective devices after they are installed in a system is based upon the conclusion or design philosophy, which we are now reevaluating, that a test of a protective device after installation is not warranted where such a test could possibly damage the component that is being protected.
Based upon the detailed consideration of the flight failure records and some very extensive "detective work" of personnel from the Lewis Research Center, which is responsible for the Centaur stage, and General Dynamics, the prime contractor, we developed the following hypothesis which we believe pinpoints the cause of the Mariner 8 failure :
(1) One of the Zener diodes which was intended to protect the inte-grated circuit in the Centaur pitch rate gyro preamplifier was either installed improperly, was faulty itself, or failed at some time during assembly, vehicle checkout or powered flight.
(2) An electrical transient sufficient to cause damage to the inte-grated circuit occurred some time after T-minus-40 minutes—when the autopilot performed properly--but before Centaur main engine start.
(3) This damaged the integrated circuit affecting the gain of the autopilot pitch channel and causing the subsequent fixed engine position.
Tests show that all subsequent mission events would be fully explained by this hypothesis.
...
To prevent a recurrence of the problem which occured on Mariner 8, we required additional thermal and vibration tests on the Centaur state autopilot and new and special test procedures to ensure that the diodes protection the integrated circuits were properly installed and operative."
Their procedure was to test each component before installation only, in order to prevent any potential damage occurring during post-installation testing. So they evidently couldn't know to scrap it and replace the damaged one. After the Mariner 8 failure, they changed this testing policy to test PCB's/components again after components are installed on the PCB.
"A diode intended to protect the system from transient voltages was thought to have been damaged during repairs/installation of the pitch amplifier's printed circuit board, something that would not have been detected through bench tests." - Wikipedia, but I'm having difficulty finding any original sources that validate or expand upon this.
P.S. I think you may have responded to the wrong parent comment
Edit: "N72-13963" was a report given to Congress in October 1971 on the failure of the Mariner 8 mission, I can't find that report but it may have more details if you're curious enough to track it down.
Edit: I found a transcript of the report given to Congress, evidently in June of 1971: https://books.google.com/books?id=7FQVAAAAIAAJ
Dr. Low: The diode was external to this integrated circuit, mounted next to it. I don't have one with me, but the diode is another component about the same size. The integrated circuit which failed is known to be sensitive to electrical transients if it is not properly protected. And electrical transients large enough to damage an unprotected integrated circuit — caused, for example, by engine start, stage separation, et. cetera — are well within the specifications of the Atlas/Centaur vehicle and are clearly possible.
Because of this, the integrated circuit was protected with Zener diodes which in effect were designed to act as "safety valves" to prevent potentially damaging electrical transients from reaching the integrated circuit itself. In accordance with standard practice in the Atlas/ Centaur vehicle — as well as in many others — these Zener diodes were checked individually prior to their installations, but were not tested thereafter. This practice of not testing protective devices after they are installed in a system is based upon the conclusion or design philosophy, which we are now reevaluating, that a test of a protective device after installation is not warranted where such a test could possibly damage the component that is being protected.
Based upon the detailed consideration of the flight failure records and some very extensive "detective work" of personnel from the Lewis Research Center, which is responsible for the Centaur stage, and General Dynamics, the prime contractor, we developed the following hypothesis which we believe pinpoints the cause of the Mariner 8 failure :
(1) One of the Zener diodes which was intended to protect the inte-grated circuit in the Centaur pitch rate gyro preamplifier was either installed improperly, was faulty itself, or failed at some time during assembly, vehicle checkout or powered flight.
(2) An electrical transient sufficient to cause damage to the inte-grated circuit occurred some time after T-minus-40 minutes—when the autopilot performed properly--but before Centaur main engine start.
(3) This damaged the integrated circuit affecting the gain of the autopilot pitch channel and causing the subsequent fixed engine position.
Tests show that all subsequent mission events would be fully explained by this hypothesis.
...
To prevent a recurrence of the problem which occured on Mariner 8, we required additional thermal and vibration tests on the Centaur state autopilot and new and special test procedures to ensure that the diodes protection the integrated circuits were properly installed and operative."
> Being that this was 1971, it was probably more straightforward to repair the botched installation/repair rather than do the whole PCB over from scratch
As an extra data point I did a contracting gig in early '90s that involved building a diagnostic system for a specialised analogue circuit board that was used in the oil industry. The boards were held in place and an automated probe system took measurements at numerous places. A constraint satisfaction algorithm then modelled failing components in an attempt to match the faulty readings, and thus locate the failed component(s). The algorithm came from a three-year Ph.D project funded by the company concerned, and I was part of a team of five that were implementing it over a six-month period.
This was considered a small price by the company, compared with the cost of throwing away a specialised board manufactured in small production runs to high tolerances. The probing system already existed, and was used by electronics experts who were attempting to perform the same diagnostic process using their own (expensive) skill and judgement.
As an extra data point I did a contracting gig in early '90s that involved building a diagnostic system for a specialised analogue circuit board that was used in the oil industry. The boards were held in place and an automated probe system took measurements at numerous places. A constraint satisfaction algorithm then modelled failing components in an attempt to match the faulty readings, and thus locate the failed component(s). The algorithm came from a three-year Ph.D project funded by the company concerned, and I was part of a team of five that were implementing it over a six-month period.
This was considered a small price by the company, compared with the cost of throwing away a specialised board manufactured in small production runs to high tolerances. The probing system already existed, and was used by electronics experts who were attempting to perform the same diagnostic process using their own (expensive) skill and judgement.
Just out of terrible curiosity, was that PCB used in downhole tools?
They were used in a monitoring / comms role at the well head if I recall correctly. Younger me just concentrated on coding the diagnostics, and as a contractor was totally insulated from actual users who might have explained.
Curious if they have to register/coordinate their intended flight path/orbits so they don't hit each other(probably very unlikely). Probably don't have onboard radar or something like that?
> probably very unlikely
Literally astronomically unlikely. If at all they might want to coordinate radio emissions.
Literally astronomically unlikely. If at all they might want to coordinate radio emissions.
“Scientists have calculated that the chances of something so patently absurd actually existing are millions to one.
But magicians have calculated that million-to-one chances crop up nine times out of ten.”
Terry Pratchett in Mort
Terry Pratchett in Mort
Sergeant Colon looked wretched. "Weeell, what if it's not a million-to-one chance?" he said.
Nobby stared at him.
"What d'you mean?" he said.
"Well, all right, last desperate million-to-one chances always work, right, no problem, but...well, it's pretty wossname, specific. I mean, isn't it?"
"You tell me," said Nobby.
"What if it's just a thousand-to-one chance?" said Colon agonizedly.
"What?"
"Anyone ever heard of a thousand-to-one shot coming up?"
Carrot looked up. "Don't be daft, Sergeant," he said. "No one ever saw a thousand-to-one chance come up. The odds against it are—" his lips moved—"millions to one."
"Yeah. Millions," agreed Nobby.
"So it'd only work if it's your actual million-to-one chance," said the sergeant.
"I suppose that's right," said Nobby.
"So 999,943-to-one, for example—" Colon began.
Carrot shook his head. "Wouldn't have a hope. No one ever said, 'It's a 999,943-to-one chance but it just might work.'"
- Guards! Guards!
Nobby stared at him.
"What d'you mean?" he said.
"Well, all right, last desperate million-to-one chances always work, right, no problem, but...well, it's pretty wossname, specific. I mean, isn't it?"
"You tell me," said Nobby.
"What if it's just a thousand-to-one chance?" said Colon agonizedly.
"What?"
"Anyone ever heard of a thousand-to-one shot coming up?"
Carrot looked up. "Don't be daft, Sergeant," he said. "No one ever saw a thousand-to-one chance come up. The odds against it are—" his lips moved—"millions to one."
"Yeah. Millions," agreed Nobby.
"So it'd only work if it's your actual million-to-one chance," said the sergeant.
"I suppose that's right," said Nobby.
"So 999,943-to-one, for example—" Colon began.
Carrot shook his head. "Wouldn't have a hope. No one ever said, 'It's a 999,943-to-one chance but it just might work.'"
- Guards! Guards!
I am curious to know this - how do various nations/ space programs ensure that the objectives of their missions are significantly different that they don't end up repeating the work/ results of other missions (from other nations)? Do all space agencies work together to determine what the focus would be (I guess not, but it'd be cool, if it were like that)!
Digital camera technology is leaps and bounds beyond the last rovers we sent. Hopefully we get top quality photos this time around.
COTS sensors don't work well in a high radiation environment.
isn't transmission bandwidth a bigger concern?
I know I'm a fuddy duddy, but I absolutely hate the naming scheme of our mars rovers - Curiosity and Perseverance are things you feel not things you do. I just hate it. Stop it NASA.
xupybd(3)