open source Mars Lander
December 30, 2008 6:39 PM
Why isn't there an open source Mars Lander? The Phoenix Lander cost a Bunch'O'Money. How cheaply could it be done without NASA involved?
The assertion that amateurs could get a working research vehicle on Mars based on the fact that some rocketry and radio hobbyists just barely managed to get an object into space 4 years ago - nearly 60 years after governments first did so - seems sort of stupid.
posted by nanojath at 7:07 PM on December 30, 2008
posted by nanojath at 7:07 PM on December 30, 2008
You're confusing "getting into space" with getting into space.
"Getting into space" requires only enough speed to reach a high altitude, and then fall down again. "Getting into space" is easy; you can do it going Mach 3.
Getting into space for realses, getting up into an orbit, is a vastly different endeavor. Not Mach 3. Mach 20+. Hugely more energy -- to get to Mach 20, you have to accelerate all the fuel you need to get you from Mach 19 to Mach 20 up to Mach 19. And you have to accelerate all the fuel you need to get from Mach 18 to Mach 19 up to Mach 18, and so on. This is why rockets on the pad are ginormous, and why rocket engines on orbital rockets are hugely powerful and why real rocket engines require hugely more engineering than toy rocket engines do.
Googling, getting to orbit requires about sixty times as much energy as a ballistic shot up to the edge of space.
Getting into interplanetary trajectories is another jump in energy above even getting into orbit.
posted by ROU_Xenophobe at 7:25 PM on December 30, 2008
"Getting into space" requires only enough speed to reach a high altitude, and then fall down again. "Getting into space" is easy; you can do it going Mach 3.
Getting into space for realses, getting up into an orbit, is a vastly different endeavor. Not Mach 3. Mach 20+. Hugely more energy -- to get to Mach 20, you have to accelerate all the fuel you need to get you from Mach 19 to Mach 20 up to Mach 19. And you have to accelerate all the fuel you need to get from Mach 18 to Mach 19 up to Mach 18, and so on. This is why rockets on the pad are ginormous, and why rocket engines on orbital rockets are hugely powerful and why real rocket engines require hugely more engineering than toy rocket engines do.
Googling, getting to orbit requires about sixty times as much energy as a ballistic shot up to the edge of space.
Getting into interplanetary trajectories is another jump in energy above even getting into orbit.
posted by ROU_Xenophobe at 7:25 PM on December 30, 2008
To the root question, open source software is based in part on the idea that the software can start off very buggy or with limited features, and can be improved quickly because new iterations are cheap.
In contrast, space software is based on the idea that iterations are expensive, and that the first try has to be feature complete and perfect. "Software assurance" concepts and such, where you translate directly from requirements to code, and provide multiple fallbacks in case there are issues. There's no particular reason to think that open source would be a good approach.
posted by smackfu at 7:34 PM on December 30, 2008
In contrast, space software is based on the idea that iterations are expensive, and that the first try has to be feature complete and perfect. "Software assurance" concepts and such, where you translate directly from requirements to code, and provide multiple fallbacks in case there are issues. There's no particular reason to think that open source would be a good approach.
posted by smackfu at 7:34 PM on December 30, 2008
Yea... just to reiterate, while the recent accomplishments of private space exploration companies have been interesting and the future of these endeavors looks promising, there is absolutely no comparison to NASA (not to mention that they aren't really open source, by definition). Money buys a lot of different things: materials, knowledge, risk-management... to just "open source" a remote-operated Mars vehicle is - and I'm sorry to put it this way - absolutely ridiculous. Private entities will probably achieve what NASA has already, but not for a long time and not without a lot of money and a lot of trusted, experienced minds working on it. Open source is no silver bullet. Would you really want an anonymous or at least un-vetted network of people doing the calculations necessary to engineer these crafts and to plan their journey? Would you honestly think it could work? And like people above me have stated, there is a vast infrastructure already in place that is vital to the success of these missions. So, in short, the reason why there is no open source Mars Lander is that it is simply not a realistic way to accomplish the task, which is much, much more complicated, expensive and fraught with risk than anyone imagines.
posted by kurtroehl at 7:52 PM on December 30, 2008
posted by kurtroehl at 7:52 PM on December 30, 2008
Yes, well -- in the UK (the homeland of DIY aka open-source), we sort of tried this, with the Beagle Mars Lander ...
posted by Susurration at 8:01 PM on December 30, 2008
posted by Susurration at 8:01 PM on December 30, 2008
I wouldn't say that it's unrealistic -- there are people who have spent more on hobbies. Hell, I just dropped a month's pay on my dogs in the past month between medical care and other items.
However, I wouldn't say that open source is the way to do it for the reasons detailed above. You might be getting some of the principles of the open source movement mixed up with the execution of the open source movement. The principles of the open source movement include "many contributors make otherwise insurmountable tasks easy" and "faster, better, cheaper". The execution that has worked best so far is "many iterations" -- something that's not easy to do with rockets, and definitely impossible to do with humans.
The 2004 ham radio launch you linked to is the equivalent of what SpaceShipOne did the same year. They passed the suborbital boundary into space -- not an orbit shot. The payload immediately came back to earth. Note that they'd been trying to do that for years and had done at least one previous "shot".
What would an independent attempt to reach Mars look like? Well, it'd probably be very slow, very small, and use cheap, well-understood parts. That execution is moot, unfortunately. The most expensive part of getting into orbit, as already mentioned, is going to be the fuel. There just isn't a more efficient way of gaining altitude right now than oxidizing (aka burning) a fuel. Fuel is heavy. Until there's a way around that (i.e. space elevator, magnetic railgun on a global scale, etc), you're not going to find any real amateur space flight.
Heck, just reaching the moon with a radio beacon is probably outside of the realm of amateur space flight right now.
posted by SpecialK at 8:20 PM on December 30, 2008
However, I wouldn't say that open source is the way to do it for the reasons detailed above. You might be getting some of the principles of the open source movement mixed up with the execution of the open source movement. The principles of the open source movement include "many contributors make otherwise insurmountable tasks easy" and "faster, better, cheaper". The execution that has worked best so far is "many iterations" -- something that's not easy to do with rockets, and definitely impossible to do with humans.
The 2004 ham radio launch you linked to is the equivalent of what SpaceShipOne did the same year. They passed the suborbital boundary into space -- not an orbit shot. The payload immediately came back to earth. Note that they'd been trying to do that for years and had done at least one previous "shot".
What would an independent attempt to reach Mars look like? Well, it'd probably be very slow, very small, and use cheap, well-understood parts. That execution is moot, unfortunately. The most expensive part of getting into orbit, as already mentioned, is going to be the fuel. There just isn't a more efficient way of gaining altitude right now than oxidizing (aka burning) a fuel. Fuel is heavy. Until there's a way around that (i.e. space elevator, magnetic railgun on a global scale, etc), you're not going to find any real amateur space flight.
Heck, just reaching the moon with a radio beacon is probably outside of the realm of amateur space flight right now.
posted by SpecialK at 8:20 PM on December 30, 2008
The real cost of a spacecraft is planning, engineering, testing, launch, mission, support, post-mission, and a slew of other things I am forgetting. Only a very small part of that is software. Even then, I think its absurd to think some OSS hackers are going to piece together the quality software needed for a space mission. Heck, we're not just talking the firmware of the craft itself but of all mission software. At this level there's really not much difference between hardware and software. The hardware and software teams work together to engineer the spacecraft.
Or are you suggesting some kind of open source hardware with the quality that matches NASA? Thats even more absurd. We dont even see that quality in the private enterprises that arent even able to achieve orbit yet!
I think its easy to sit back and see how well linux works and think its applicable to everything, but its not. The quality of linux, and other mature OSS projects, were terrible in the beginning. This methodology wont work with a 200 million dollar space mission.
NASA engineers really have no need for a lot of the OSS stuff out there. OSS culture still revolves around the desktop PC. Spacecraft need real-time OSs and very specialized hardware. I could see taking an existing high quality real-time OSS OS (are there any?) and modifying it for use, but even that seems like an edge case to me.
Im sure there's some OSS stuff used for mission control desktops and servers, but as a cost saving measure for spacecraft engineering? Doubtful.
This is not to dismiss the DIY approach. I believe some college students have had mini satellites launched into a low orbit before, but only when the government is footing the launch bill. OSS has also saved cost in the word of academia, especially when you price out commercial unix systems.
Lastly, openness and collobartion pre-date OSS by quite a bit. A lot of these projects are tied to industry or universities. There are protocols in place to make sure everyone is speaking the same language and using the same tools. NASA is probably the most open government organization.
posted by damn dirty ape at 9:14 PM on December 30, 2008
Or are you suggesting some kind of open source hardware with the quality that matches NASA? Thats even more absurd. We dont even see that quality in the private enterprises that arent even able to achieve orbit yet!
I think its easy to sit back and see how well linux works and think its applicable to everything, but its not. The quality of linux, and other mature OSS projects, were terrible in the beginning. This methodology wont work with a 200 million dollar space mission.
NASA engineers really have no need for a lot of the OSS stuff out there. OSS culture still revolves around the desktop PC. Spacecraft need real-time OSs and very specialized hardware. I could see taking an existing high quality real-time OSS OS (are there any?) and modifying it for use, but even that seems like an edge case to me.
Im sure there's some OSS stuff used for mission control desktops and servers, but as a cost saving measure for spacecraft engineering? Doubtful.
This is not to dismiss the DIY approach. I believe some college students have had mini satellites launched into a low orbit before, but only when the government is footing the launch bill. OSS has also saved cost in the word of academia, especially when you price out commercial unix systems.
Lastly, openness and collobartion pre-date OSS by quite a bit. A lot of these projects are tied to industry or universities. There are protocols in place to make sure everyone is speaking the same language and using the same tools. NASA is probably the most open government organization.
posted by damn dirty ape at 9:14 PM on December 30, 2008
India just got to the moon for ~78m dollars
posted by bottlebrushtree at 9:35 PM on December 30, 2008
posted by bottlebrushtree at 9:35 PM on December 30, 2008
FWIW, the mars landers use an OS called VxWorks. Could this be replaced by an open source application like RTEMS? Possibly, but the hard part is getting to mars in the first place.
Also, an interesting article on the ESA's Beagle 2 mission control running on SUSE. For unrelated reasons Beagle-2 was a failure.
posted by damn dirty ape at 9:42 PM on December 30, 2008
Also, an interesting article on the ESA's Beagle 2 mission control running on SUSE. For unrelated reasons Beagle-2 was a failure.
posted by damn dirty ape at 9:42 PM on December 30, 2008
Earth escape velocity is 11.186 km/s. The speed of sound is 343 m/s.
So escape velocity is greater than Mach 32.
posted by Chocolate Pickle at 9:49 PM on December 30, 2008
So escape velocity is greater than Mach 32.
posted by Chocolate Pickle at 9:49 PM on December 30, 2008
The government is still involved. Who would pay for this? Lando Calrissian, looking for diamonds on Mars? Launches are big business and cost a lot of money. Only the government would pay that sort of money, because there's no business return on investment. It is a public good.
The commericial space outfits are doing it for money--space tourism, people paying. There's no Bond Villan type shelling out millions for a Mars probe.
Plus you could throw money at the programming side and it still wouldn't make a dent. Its the cost of lifting it into interstellar space which is what drives it. NASA, ESA and the Russian government are the only organizations capable of this.
posted by Ironmouth at 10:04 PM on December 30, 2008
The commericial space outfits are doing it for money--space tourism, people paying. There's no Bond Villan type shelling out millions for a Mars probe.
Plus you could throw money at the programming side and it still wouldn't make a dent. Its the cost of lifting it into interstellar space which is what drives it. NASA, ESA and the Russian government are the only organizations capable of this.
posted by Ironmouth at 10:04 PM on December 30, 2008
Yeah, the software cost is the least of your worries.
posted by delmoi at 11:38 PM on December 30, 2008
posted by delmoi at 11:38 PM on December 30, 2008
It's happening. The civilian/amateur/enthusiest Planetary Society (website) designed and built themselves a solar-sail spacecraft, and raised sufficient funds to purchase a launch into space for it from the Russian space agency. Unfortunately, the launch was not successful and the spacecraft was lost.
The risk of a failed launch is something faced by the corporate and government owners of every satellite and spacecraft that goes up (or not, as the case might be), and there is generally sufficient finance in place to weather the risk, but in the case of this (almost) bake-sale funded spacecraft, they only had enough for one shot, and the loss was catastrophic.
posted by -harlequin- at 1:13 AM on December 31, 2008
The risk of a failed launch is something faced by the corporate and government owners of every satellite and spacecraft that goes up (or not, as the case might be), and there is generally sufficient finance in place to weather the risk, but in the case of this (almost) bake-sale funded spacecraft, they only had enough for one shot, and the loss was catastrophic.
posted by -harlequin- at 1:13 AM on December 31, 2008
Damn dirty ape's comment,
"The real cost of a spacecraft is planning, engineering, testing, launch, mission, support, post-mission, and a slew of other things I am forgetting."
reminded me of a book I recently read, John McPhee's The Deltoid Pumpkin Seed. It's about a company - or, perhaps more accurately, a couple of visionary directors and a handful of mechanics, engineers, and test pilots - trying to build a fleet of dirigibles, starting with prototypes. I found it a surprisingly gripping read, and it has some lovely turns of phrase (it's John McPhee, after all). It might give you some idea of what a project like that could look like, and what kinds of challenges it might face. (Funding, mostly, and then execution.)
posted by kristi at 10:35 AM on December 31, 2008
"The real cost of a spacecraft is planning, engineering, testing, launch, mission, support, post-mission, and a slew of other things I am forgetting."
reminded me of a book I recently read, John McPhee's The Deltoid Pumpkin Seed. It's about a company - or, perhaps more accurately, a couple of visionary directors and a handful of mechanics, engineers, and test pilots - trying to build a fleet of dirigibles, starting with prototypes. I found it a surprisingly gripping read, and it has some lovely turns of phrase (it's John McPhee, after all). It might give you some idea of what a project like that could look like, and what kinds of challenges it might face. (Funding, mostly, and then execution.)
posted by kristi at 10:35 AM on December 31, 2008
Actually, much of aeronautical software is open source. Just not open to you. For various reasons, like "aiming a rocket at the US is just like aiming a rocket at mars". Also "because people are willing to pay for the source".
QNX, VxWorks and other "real time embedded" operating systems have source available if you have the money available. VxWorks for example, is what replaced Linux in the newer Linksys WRT54G routers; they were able to cut RAM requirements in half as a result. As others have pointed out, software is the least expensive of concerns (although still capable of exploding millions of dollars in equipment). Currently open source hardware is almost non-existent. Space radiation resistant hardware? Who could afford that?
Finally, there are some parts of NASA operation that are open source.
posted by pwnguin at 3:38 PM on December 31, 2008
QNX, VxWorks and other "real time embedded" operating systems have source available if you have the money available. VxWorks for example, is what replaced Linux in the newer Linksys WRT54G routers; they were able to cut RAM requirements in half as a result. As others have pointed out, software is the least expensive of concerns (although still capable of exploding millions of dollars in equipment). Currently open source hardware is almost non-existent. Space radiation resistant hardware? Who could afford that?
Finally, there are some parts of NASA operation that are open source.
posted by pwnguin at 3:38 PM on December 31, 2008
At least 10 satellites built by amateurs, hobbyists, and students are currently functioning, out of dozens launched over the last 40 years. The international nonprofit that coordinates these projects is AMSAT.
Just like open-source software, most of the amateur satellite builders are university students working for class credit, or professional engineers who volunteer their spare time.
Getting into orbit is indeed expensive -- which is why AMSAT satellites usually ride on a converted missile (which can carry just a few kilos to orbit), fill the excess capacity of a commercial launch, or serve as test payloads for unproven new rockets. Sometimes these flights are paid for by governments, other times they are funded by donations. Satellites often sit on the ground for years, waiting for such an opening. The typical cost is just a few million dollars.
The German chapter of AMSAT is currently raising funds for a Mars mission. The plans are sketchy right now, but this is a group of hardcore hobbyists who have successfully put seven satellites into orbit on a shoestring, so I wouldn't discount them too soon.
posted by miyabo at 2:06 PM on January 2, 2009
Just like open-source software, most of the amateur satellite builders are university students working for class credit, or professional engineers who volunteer their spare time.
Getting into orbit is indeed expensive -- which is why AMSAT satellites usually ride on a converted missile (which can carry just a few kilos to orbit), fill the excess capacity of a commercial launch, or serve as test payloads for unproven new rockets. Sometimes these flights are paid for by governments, other times they are funded by donations. Satellites often sit on the ground for years, waiting for such an opening. The typical cost is just a few million dollars.
The German chapter of AMSAT is currently raising funds for a Mars mission. The plans are sketchy right now, but this is a group of hardcore hobbyists who have successfully put seven satellites into orbit on a shoestring, so I wouldn't discount them too soon.
posted by miyabo at 2:06 PM on January 2, 2009
This thread is closed to new comments.
Receiving weak radio signals from Mars requires big dish antennas. NASA has a world-wide network of such receivers, and without them it would be very hard to receive any telemetry from your hypothetical lander.
There are a number of other issues involved that make it nontrivial, but those two are enough.
posted by Chocolate Pickle at 6:51 PM on December 30, 2008