Slashdot Mirror
Schiaparelli Mars Probe's Parachute 'Jettisoned Too Early', Whereabouts Still Unknown (bbc.com)
An anonymous reader writes: Europe's Schiaparelli lander did not behave as expected as it headed down to the surface of Mars on Wednesday. Telemetry data recovered from the probe during its descent indicates that its parachute was jettisoned too early. The rockets it was supposed to use to bring itself to a standstill just above the ground also appeared to fire for too short a time. The European Space Agency (Esa) has not yet conceded that the lander crashed but the mood is not positive. Experts will continue to analyse the data and they may also try to call out to Schiaparelli in the blind hope that it is actually sitting on the Red Planet intact. In addition, the Americans will use one of their satellites at Mars to image the targeted landing zone to see if they can detect any hardware. Although, the chances are slim because the probe is small. For the moment, all Esa has to work with is the relatively large volume of engineering data Schiaparelli managed to transmit back to the "mothership" that dropped it off at Mars - the Trace Gas Orbiter.
I suspect something happened to the parachute just before jettisoning - e.g. rope breaking/twisting, chute tearing/burning that caused the lander to tumble, and so the rockets switched off as they weren't pointing against the forces on the lander that the accelerometers were reading.
Look at JPL's Mars parachute test in 2014 that ripped the parachute https://www.youtube.com/watch?....
I was very disappointed and angry at the ESA Press Conference this morning. Last night when they suspected (knew) [There is no way this landing was designed to not return a success signal immediately] the Lander had crashed they silenced everyone and announced a press conference at 10 AM the next morning.
At the Press Conference they emphasised the success of the orbiter and mentioned NOTHING at all about data from the lander. They left that all to questions from the Press. Basically all questions from the Press were about the lander and the data (and they were good questions - no stupid questions came), and they drip fed a piece of info at a time to the journalists.
I believe the suits at ESA were in damage control because they are scared about losing funding for the 2020 lander so they mentioned NOTHING about the crashed lander, so that when politicians check on the press releases/conferences in months to come there is NO info on the crash, but in a few days the world will know anyway, especially if NASA gets a photo of the impact and debris.
They did not make any statements at all (e.g. yes we got data from the Lander, the rockets fired for only 3-4 seconds, something went wrong with the parachute and we suspect a very hard landing) in a controlled and orderly way, they forced the journalists to extract it from them relunctantly.
I was super disappointed about scientists playing politics and covering up what they obviously knew the audience wanted to know. It was sickening.
And, yes, I live in Europe and yes, I want my tax Euros to fund the 2020 Lander, but I'm angry at scientists playing politics and ignoring the audience who wanted to know what happened last night and they deliberately said nothing.
Someone should go to jail for a very long time as soon as we figure out exactly what was screwed up.
Really. Criminal conviction, huh? Programmer in prison? Are you even listening to yourself?
Would it be the Programmer? Or the Tester? Or the Project Manager that signed off on it? Or should it be the entire team?
Hillary Clinton. Trump says we should blame her for everything that has gone wrong - ever. :-)
It must have been something you assimilated. . . .
Quite possibly none of the above; from the description, I think that it's more likely some sort of sensor issue. It used a doppler radar altimeter/velocimeter to estimate its position, with 1 antenna is dedicated to range (points straight down, direct measurement) and 3 to X/Y/Z velocity (angled outward, used to estimate how the landscape is moving with respect to the craft). There's also accelerometers onboard. I'm not sure what sort of priority is given to what data.
A program is only as good as the inputs it receives. It seems to me that it thought it was going "low and slow". I mean, technically it could be a software issue, there could be some sort of "unit conversion" bug or some sort of mistaken sequence specifications or the like. But if I had to guess, I'd go with a sensor data problem rather than software.
"99 dead duelists of Dios on the wall. 99 dead duelists of Dios! Take one's ring, pass it around..."
Real engineers go to jail when they fuck up.
People who pretend to be engineers go to jail. Did you actually read your link? It's nothing but people with forged documents and other fraudulent acts. Nothing about sending people to jail for honest errors where nobody died.
Maybe a little accountability for you "coding is an art" folks would be a good thing?
If someone is willing to pay for the proper quality control structures then fine. Most software engineering quality control is severely budget limited. There are folks out there who know how to write incredibly robust software but doing that isn't cheap and it isn't merely a matter of throwing money at the problem either. It's not a secret how to do it but it isn't cheap and it isn't easy. If you want people to do a better job then you need to give them the resources and organizational structure necessary to make it happen.
Answer this. Would you do a job where you could go to jail for making a error in a calculation? Especially if no one was injured?
That last sentence, if you assume loss of signal corresponds to impact with the ground, suggests de-orbit velocity relative to the ground was much higher than expected. The early parachute release may have been the culprit. Or the probe entered the atmosphere at too steep an angle (which could also explain the early parachute release - the probe would've entered higher density atmosphere more quickly thus increasing aerodynamic load on the chute to the point at which it failed). The burn probably began at a higher velocity than it was designed for.
If we're speculating, my guess would be the higher velocity when the retro-rockets were fired caused greater instability - aerodynamic forces caused the probe to rock more than expected. The parachute's purpose isn't just to slow the craft down; it also keeps the craft's orientation stable during this period of higher aerodynamic forces. Without it, drag on tiny asymmetries on the front of the craft can result in large turning moments. With a parachute attached, these moments are countered by the righting moment the parachute imparts on the rear of the craft every time it deviates from the proper orientation. Without the parachute, the craft can experience large oscillations or even flip due to these drag-induced turning moments. The large amplitude and higher frequency of the resulting oscillations could've exceeded what the rocket control software was designed to handle, and it shut off prematurely when it exceeded some threshold programmed into the software.