The Phoenix rises.

August 5, 2007 7:45 PM

NASA's Phoenix probe launched Saturday from Cape Canaveral, destination Mars. Its mission is to investigate polar ice. This probe is unique for a couple of reasons: first, it will face a traditional parachute-and-retro-rockets landing, unlike previous endeavors. Second, it will be landing far north of any previous mission. Previous Mars missions have had mixed success, with only about half successfully making it to their destination. It is scheduled to land in May, 2008.
posted by backseatpilot (16 comments total)

[This is good]
posted by dirtynumbangelboy at 8:16 PM on August 5, 2007

robots exploring space is good
posted by b1tr0t at 8:23 PM on August 5, 2007

Today's Astronomy picture of the Day is apposite.
posted by nowonmai at 8:41 PM on August 5, 2007

It's surprising that automated retro rockets work at all. And that they worked in the 70's is miraculous.
posted by smackfu at 9:24 PM on August 5, 2007

A couple of the previous probes which didn't make it were lost because of mistakes, mostly because quality control was given short shrift under the "better, faster, cheaper" policy of the NASA director at the time.

Let's hope that this time they decided that "better" was more important than "cheaper". A probe which doesn't arrive doesn't do anyone any good, no matter how fast or cheap it was to produce.
posted by Steven C. Den Beste at 9:36 PM on August 5, 2007

I was late for something so I didn't get a screenshot, but CNN had the fantastic headline

NASA probe
begins 422 mile
journey to Mars

Steven, don't forget that one of the motivations for "faster, better, cheaper" was the loss of the $1 billion Mars Observer mission.

The very successful f/b/c idea was distributing redundancy across several missions instead of trying to have one that was very redundant in terms of communications options or instrumentation, but wholly dependent on non-redundant launch vehicles. As a result we have multiple orbiters and multiple landers at Mars today.
posted by dhartung at 9:53 PM on August 5, 2007 [1 favorite]

I'm a close friend of one of the people who built this. The last couple of months have been very busy for her. If her schedule is anything to go by, if this one fails it won't be due to a lack of testing (knock wood.)
posted by kyrademon at 10:00 PM on August 5, 2007

Dan, the problem with FBC is "pick any two".

There was a probe that was supposed to land near the south pole. It was lost, and it turns out that there was a systemic problem that was missed because it wasn't tested as a system. There were sensors on the legs which indicated that they were under pressure and extended. This was supposed to detect that the probe had landed, and upon seeing it, the computer would cut the rocket.

The problem was that just before insertion, the legs were spread out, and they overextended slightly, because of the spring. That triggered that sensor briefly and its output was latched, so the software didn't fire the rocket at all.

The Wikipedia article about Mars Observer notwithstanding, what I heard is that the reason it was lost is that someone entered a value into a computer in miles instead of in kilometers. It was an altitude to be used during insertion, and as a result the computer plotted the braking pass through the Martian atmosphere too low.

Those are the kinds of things that happen when you're trying to cut costs.

The Hubble was nearly rendered useless because of unwillingness to perform an acceptance test of the mirror. It turns out that the Hubble mirror is one of the best and most true mirrors ever ground -- but it wasn't ground to the shape they thought it was being ground to. As a result, it didn't match the rest of the optics.

That could have been detected on the ground -- but NASA didn't think it needed to perform the test. If they had, they could have reworked all the rest of the optics to match the mirror's real shape. As it was, they only detected the problem on first light, after the scope was in orbit. Then they had to design a retrofit package to go into one of the instrument bays to fix it. That cost us a couple of years of observation time, and a lot of money for an unexpected shuttle mission.

In practice, "Faster, better, cheaper" has meant "There's never time to do it right, but there's always time to do it over." Only when it comes to Mars missions, you don't get do-overs.
posted by Steven C. Den Beste at 10:20 PM on August 5, 2007

Dig this groovy liftoff shot!
posted by rob511 at 10:59 PM on August 5, 2007

We were supposed to have people farming on Mars by the time I was in my 40s, damn it, but I guess I'm happy that at least we're still sending bots up there. Go Phoenix!
posted by stavrosthewonderchicken at 12:27 AM on August 6, 2007

That's just great, now the Decepticons will know we're here.
posted by Brandon Blatcher at 2:41 AM on August 6, 2007

The Wikipedia article about Mars Observer notwithstanding, what I heard is that the reason it was lost is that someone entered a value into a computer in miles instead of in kilometers.

It was the Mars Climate Orbiter that ended up lithobraking because of the unit swap, not the MPL. The swap wasn't miles/kilometers, it was force, input as pounds, when the system was expecting newtons, which meant the input value was about 4.5 times larger than intended. The reason the problem made it to flight was that the for the first time, there wasn't an end-to-end test of the ground and flight control software for the angular control systems (a combination of reaction wheels and thrusters.)

So, the insertion burn ended up on the wrong vector. Lithobraking happens.

The Mars Observer went dark days before the orbital insertion burn. Best guess was an explosion when they started to pressurize the hydrazine lines, but we're guessing, really -- it went dead and stayed dead. It could well be in Mars Orbit if it was merely a comm failure (though radar should have seen it by now, given the number of active missions at Mars, and the scans done to check orbital insertions and not.)

Mostly likely, Mars Observer is in a solar orbit.
posted by eriko at 4:58 AM on August 6, 2007 [1 favorite]

Note that this was a bad streak for *everybody.* We lost the Mars Observer, Mars Polar Lander, and Mars Climate Observer missions, and Deep Space 2 went "splat" as well. Japan lost the Nozumi (Hope) mission. Russia launched the very ambitious Mars 96 orbiter, alas, the fourth stage firing resulting in aerobraking in the Earth's atmosphere. The EU lost Beagle 2 on landing, though the Mars Express orbiter carrying it was successful.

They weren't all losers, it just seemsed that way. Mars Pathfinder worked, as did Mars Global Surveyor, which finished the primary and two extended missions before failure. Mars Express, as mentioned, is still in operation. Since the Beagle 2 splat, we've had a good record -- Mars Odyssey and Mars Reconnaissance orbiters both worked flawlessly, and while Beagle 2 failed, Mars Express has been doing a great job, and of course, the amazing success of MER-A and MER-B.

Better, Faster, Cheaper did have some problems, but it did lead to a change in mindset for the good -- typically because most of the B/F/C failures were stupid ones. The missions we have had for the last decade really haven't been B/F/C, but they haven't been the huge probes of earlier years, with Cassini being the last of the Huge Probes. (Part of that reason is the loss of the Huge Boosters to fly them, but most is the cost.)

So, now we get smaller missions that are working very well, and can be put together far quicker than the Huge Probes. Now, we can get mission built and launched in the third of the time. B/F/C may have failed, but the mindset change didn't, and the result was such things as MER, New Horizons (built it fast, or you *can't* launch it...) and now Dawn.
posted by eriko at 5:13 AM on August 6, 2007 [1 favorite]

I was surprised to learn that of 15 attempts to land on Mars, only 5 have succeeded. With all the imagery over the years, you would have thought there was more.
posted by fluffycreature at 9:55 AM on August 6, 2007

eriko 1, SCDB 0 .

Nervously watching MER power levels ...
posted by intermod at 7:27 PM on August 6, 2007

Dan, the problem with FBC is "pick any two".

Steven, don't confuse an idiom with a scientific law.

I spent plenty of time on sci.space.* watching the FBC detractors say exactly that over and over, while FBC and successors got mission after mission into orbit or onto ground. Yes, some were lost. But that clearly wasn't something that FBC introduced.

"Cheaper", especially, but all three of the criteria are not either/or propositions. They're sliding scales. There's probably a lower bound for the cost of a successful Mars mission, but certainly not an upper one for an unsuccesful mission, either. Decrease cost, increase risk. Maybe you decrease cost, and introduce some cross-mission savings such as reusable equipment, personnel, and so forth, that the increased risk is worth it. What's "worth it"? That's the subjective part of the question.

Those are the kinds of things that happen when you're trying to cut costs.

My point was that MO cost $1 billion and was in fact in its excess the disappointment that led to FBC approaches gaining momentum. Whether its failure was in fact a result of some cost-cutting somewhere isn't likely to ever be known. I'm not sure what NASA's final report on the matter says, but it's clear that after MO, FBC proponents had the upper hand. If you're going to lose a fair number of your missions, no matter how much you spend or how careful you are, the obvious answer is to send more.

Only when it comes to Mars missions, you don't get do-overs.

If you can have four probes for the price of one, you do.
posted by dhartung at 9:15 AM on August 7, 2007 [1 favorite]