Slashdot Mirror
Beagle 2 Failure Analyzed
InsomniaCity writes "An inquiry into the loss of the Beagle 2 Mars probe in December will criticize the management of the project and the testing of the lander, says the BBC. Following the loss, the European Space Agency (Esa) and the British National Space Centre established a Commission of Inquiry, that are now recommending 19 things we need to remember for the future, from project management and fund raising, to high altitude testing of the parachute system. The commission, however, did not pinpoint any particular technical failure."
With how fast it was sent through production into use, it could have been something as simple as faulty programming.
Maybe someone told it to use it's parachute at 5m instead of 5km? The world may never know.
The cost of Beagle 2 varies (depending on who you ask) between 25 and 35 million pounds. Let's take an average of 30 million. The cost of the US Mars Rovers was 800 million for 2 (with savings on each because there were 2 of them). Right there is why Beagle 2 failed. Any failures in management are going to be mere perturbations on a delta-function graph - they had the best available technology, science, and equipment for the costs they could afford.
It's interesting to note that Manchester United paid 30 million for Rio Ferdinand from Leeds United (this is English Premier League Football, for those not UK based) which sort of sums up the UK attitude to space travel. We pay roughly equivalent amounts to move a footballer about 65 miles as we do to send a robot explorer to a different planet in search of life....
I think it all starts at a very early age. Sport is instinctively popular amongst kids and remains popular amongst adults. Science for kids is boring and dull (apart from Chemistry where once in a blue moon you get to blow something up). There are tables to learn, maths equations to solve, rules and laws to learn by rote. None of this is fun.
As kids become adults, they keep their inhibitions about science
The case for the prosecution of "boring science at school" rests, M'lud.
Simon.
Physicists get Hadrons!
They are recommending 19 things we need to remember for the future, including "testing of the parachute system".
Am I the only one who can't believe they didn't think of this before?
Do you or your partner snore? - Visit www.snoring.com.au
Which is a more popular topic in a bar (or anywhere, really), the search for the Higgs Boson, or 'Who will win the league'?
When it's his turn to buy the drinks, the search for Higgs Boson wins hands down.
"Accept that some days you are the pigeon, and some days you are the statue." - David Brent, Wernham Hogg
If they'd have named it Siamese, I'm sure it would have landed right.
BigFiber.net
A couple of geeks from slashdot should have checked their documentation before they even built Beagle II ... They may have succeeded :)
Obviously, they should have named it Slashdot I ...
Never underestimate the power of idiots in large groups
I realize this is not an American probe, but the American example of Metric\American system is a prime example of things that generally go wrong on these type of missons. Generally simple, easy to avoid things are the prime kickers. While hind site is 20/20, I think some increased quality control would be quite usefull.
umm well the Mars Rover did arive a weak after beagle... sabotage I tell you
the rolls royce engines are some of the best plus they have lots of other things.
I am a bit amazed that they tested everything individually but didn't do an end to end testing.
I realize that they were on a tight schedule. I wonder if they performed environmental testing under extreme conditions? The article didn't mention it but it is really important. Especially since I doubt the mars atmospere resembles the weather in Great Britain.
I think it would be cool if they gave it another go.
Karma, We don't need no stinkin' karma!
It's ESA not Esa, do U write Usa???, michael: -1 point.
The speed at which new technologies are rushed from design through to market is a concern.
Admittedly technology that may have fatal consequences (like aircraft, flying machines and drugs)is usually subject to more regimented testing before release, but there is no shortage of products without such controls.
How many of us end-users should have been called beta-testers instead (cough...mobile phones...operating systems etc.)?
Unfortunately the rush to get the product to market before competition meant there wasn't enough time (or interest) in testing it. Was that the problem here, with the Mars Rover on the way at the same time as the Beagle?
Do you or your partner snore? - Visit www.snoring.com.au
[joke]I think the point is clear, Bush saw that the British were going to get there first. So he took it down.[/joke]
Space missions are risky and expensive. You can spend lots of extra money and have the mission fail anyway. And there's a danger of it getting cancelled altogether if you spend too long testing.
I'm not so sure that funding is the sole source of ESA's woes. Of course, when you're only budgeting 1/30th the amount as your competitors, things aren't always going to work out.
But consider the fact that ESA was founded in 1975 (May 30, 1975, to be exact. We're almost at the anniversary...) By this time, the USA had already tested numerous rockets, put men into space, and had landed men on the moon six years prior. The orbiter was already in development and the first "space shuttle" would be deployed less than a year later.
ESA has a lot of catching up to do, even now. I don't doubt that NASA and ESA share a lot of data, but for the time being, NASA has a great deal of experience over any competing space agency. Funding is not the only problem that ESA will have to overcome before it can be seen as NASA's contemporary.
--
Rate Naked People (Not work-safe)
it would never meet the deadlines because of all the new features being put into it constantly.
...the British scientific information ministers finally admitted it failed?? I thought they were holding out hope until the 2038 rollover.
I hope no-one blames Professor Pillinger for this. He did a great job under the circumstances and he deserves a chance to try again - this time with funding and commitement from ESA.
The Mosquito was actually a very successful plane. The engineers had the sense to use a highly sophisticated composite material using hollow fibres for stiffness (otherwise known as wood).
They chose wood, not to impress the ignorant, but because it was the best material for the job in the circumstances - that is largly what engineering is about.
I remember reading a few years ago about the "new" approach to space exploration. Instead of sending less probes, they (the space agencies) would be sending more, cheaper probes. The idea being that yes, there would be a higher proportion of failures, but when offset against the increased number of missions overall, we'd end up with a higher (number of successful missions) / (total expenditure across all missions).
A similar idea crops up in the manned versus unmanned debate - "unmanned exploration is cheaper because amongst other things, you don't have to be as sure the spacecraft won't fail because there's no human life at stake."
We've now got our numerous, cheaper (Beagle cost 50 million pounds), unmanned missions. But when half of them fail (der!), people get into a kink!
Next time, don't program it to land in a crater. NASA was lucky that their similarly boneheaded failure to include filesystem cleanup code in their Mars landers was repairable.
The software was actually built by LogicaCMG. At work we received a christmas card from them (i.e. before it was known that the probe had failed), saying this:
"LogicaCMG delivered the mission-critical software that controls Beagle 2 during the hazardous ride through the Martian atmosphere, releasing the heat shield and deploying parachutes and gas-filled air bags, slowing Beagle 2 down from its 14,000 mph/22,530 kph approach velocity to a safe landing on the surface of Mars"
Or maybe not - but thanks for the card anyway ;-)
I wouldn't call Beagle 2 a failure. It didn't accomplish it's scientific goals but it proved that a small group of people with all the odds against them can produce a high quality spacecraft (and it was a high quality piece of kit, excluding perhaps the parachute) and get it to another world.
And it did get to Mars! Sure, it landed much like a bowl of petunias falling from several miles would -- but the fact that it flew at all was the amazing thing. Keep an eye out for the BBC documentary on the whole mission to get an idea of what I mean.
My message to the Beagle 2 team: It's difficult getting to Mars, and for your first attempt you did really well. Better luck next time!
I saw one of the beagle-2 flight spares last week. It was displayed and discussed at the university department where I work.
One suggestion for the loss of the craft was that the barometric altimeter, which was to deploy the parachute, was fooled by an unseasonal sand storm in the Martian atmosphere. The altimeter had to trigger the chute quite late in the descent, and the low pressure associated with the storm may have inhibited the deployment until the craft hit the ground.
Since Beagle had no engines, it couldn't go into parking orbit until the storm went away.
Wouldn't surprise me at all if bad management led to the failure. I'd never grasped the value of good management until my current job, which lacks it. Not that I'm any kind of great programmer, but I always felt before that the burden of success was mine, 100%, when in fact it's a lot less than that. The people making the higher level decisions (resource/time allocation, features, scope, requirements) are the ones who can really fuck things up. Heh, that seems pretty trite now that I've typed it...
The problems decribed here are endemic to the way what little money is available for British scientific research is distributed.
I work for the Earth Sciences department at Oxford University, one of the very best funded Universities in the UK, yet much of the time which should be spent on research by lecturers and postdoctural research staff is tied up with the beaurocracy of funding. Not only this but the funds available to keep the departments running, ie. the infrastructural costs, are going down year on year.
I feel for Professor Pillinger. He did the best job of getting funding he could. It's highly unlikely that he would have be able to get more managerial help from anyone in the current circumstances and the only person who could have publicised the whole thing was himself.
If the research council and funding bodies are anything like NERC, they only want research which already knows the results (ie. pointless) and is preferably one of the fashionable subjects (currently climate change and the environment).
Please note that I am speaking on behalf of myself and not in any way on behalf of the University of Oxford or the Department of Earth Sciences. All of the opinions expressed are my own and do not necessarily reflect those of any group within the University of any policy thereof.
Agrajag: "Oh no, not again!"
If they had thought of testing the parachute before they would have considered the idea that it might fail, and planed for time to build one that works. There is no point in doing a test if you can't make use of the results. They wasted time and money pretending to test the system.
Yes I understand there were time pressures. I don't know what they could have done differently to make it work.
Well,
... or five digits on your measure meter. Level 1 is the lowest and level 5 the highest. On level 5 are world wide about 10 to 15 software development departments.
... with about 3000 lines of issues per project. Neither did they follow a decent way to gather requirements ... no way of testing ... they even used a proprietary programming language ... I could continue :D
/.) suggested that the atmosphere pressure varies in a far broader range than we assumed so far. If you have a lander scheduled to open its parashutes at pressure X and you asume you have a good enough altitude at that pressure but in fact the weather conditions are totally different .... you crash land.
....
CMM is *known* just like smeltery and pottery is *known*.
CMM is a kind of measure how able you are to do software or other engineering work. So *knowing* it is not enough, you have to apply it.
CMM has 5 levels
Even on level 3 are only about 35% of all software engineering departments(companies).
Basicly: the majority of development labs don't use any advanced software engineering practice. Example is my curent contract, a company with 40,000 employees and about 1000 software engineers and about 2000 external consultants/software engineers, they are on CMM level 1.
Before I started to help them they not even had a revision control system. Issue tracking was done with an Excel sheet
Regarding Beagel 2: I guess the Beagle project was well aware of their quallity or lack there off. However if you like to catch a train the most important thing is to be in time at the train station. The ESA Mars Express had no chance to wait for Beagel 2. There was a launch window depending on the positions of the planets during that time in the solar system. The Beagel team did the best they could to get a probe ready to be fitted into the Mars Express probe and to join the trip.
Note: Mars is still a not very well explored/understood world. Recent informations (noted here on
I would not wonder if Beagel 2 worked perfectly but crashed because of unexpected circumstances. Roughly 50% of all Mars landers crashed
angel'o'sphere
Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
At least they realize that tracking it during decent is something they should do from now on.
File under 'M' for 'Manic ranting'
I disagree, and it bothers me seing space agencies around the world trashed all the time. For example, NASA is a common punching bag for the Shuttle. Its predicted 18 million dollar launch cost, while looking great on paper, quickly swelled to 500 million per launch. Well, guess what? The Shuttle is pretty much par for the course. Almost all rocket programs have started out with grandiose dreams, and failed. Rocket launch costs, while fluctuating a lot, have remained relatively stable since the 1960s.
I was debating with someone recently who kept insulting the shuttle, and referring to Ariane. I then showed the person how much of a disaster the Ariane program has been as well - Ariane 5 having three failures in 18 paid launches, the cost overruns driving their price up (not as high as the shuttle, but still not that great), the bailout of Arianespace, the cancellation of the ESC-B upper stage, etc. The person's response? They picked another rocket system to use as their champion, ignoring the fact that *it* had its own problems too.
To make it worse, many of the people who trash space agencies treat ameteurs as if they're the ray of hope for the future. The ameteur rocket industry has been one failure after another, and has eaten enough dollars to fund some serious development at real space agencies. They're about to start getting their first major successes - and while they too have some very good people working for them, well, "Whoopee". When they've gotten several thousand designs into space hundreds or even thousands of times each, give me a call.
We've had some truly brilliant people working at places like NASA, the ESA, etc., who have achieved incredible tasks. And while one may blame the management, guess what? Decisions have to be made. I heard someone the other day criticizing NASA for embarking on the Shuttle project and treating it as junk, while glorifying the never-made Sea Dragon. Well, how on earth was NASA supposed to predict that the Shuttle's costs would increase so dramatically due to technological problems not yet discovered? What makes one think that a rocket, "built like a ship", would have *less* technical problems? In fact, when SEALAR was built based on the Sea Dragon design, its performance figures were horribly downgraded and even still it ended up with serious structural failures that led to its cancellation. And the shuttle's costs aren't actually as bad, comparitively, as many people think - ~20,000$/kg, while the cheapest launches out there, using the latest tech, are ~10,000$/kg and are not man-capable.
So give them a break, people. They're got some of the really intelligent people working very hard on an *incredibly* difficult task.
Musk needs a safer hobby than Twitter. Fire juggling? Cage fighting? Solo hot air balloon trips?
Being that it was their first probe, perhaps they should have cut down on the number of gizmos so that they had money and weight to have better landing systems.
Rather than include a dozen experiments, maybe just have 2 or 3. For example, just focus on detecting life instead of x-ray spectrometers etc.
Table-ized A.I.