Bad Software Blamed For $1.2 Billion Space Failure

http://www.usatoday.com/news/ndsmon13.htm

By Jack Williams USATODAY.com 5-11-99

CAPE CANAVERAL, Fla. (Reuters) - Corrupted computer software loaded into a Titan 4 rocket is being blamed for wrecking a $1.2 billion military space mission, the most costly in a string of six U.S. launch failures, a respected trade magazine reported Monday.

A Centaur upper stage booster on the Lockheed Martin Titan 4 rocket veered off course about nine minutes after blasting off from Cape Canaveral on April 30, leaving a sophisticated $800 million military communications satellite in the wrong orbit.

``The Centaur upper stage was launched carrying an inaccurate software load from Lockheed Martin that went undetected in the company's software verification process,'' the May 10 edition of Aviation Week and Space Technology reported.

After starting to malfunction, the incorrectly programmed booster went haywire, firing its twin engines at the wrong times and releasing its costly cargo three hours early into an orbit thousands of miles too low.

Workers at the Lockheed Martin, Littleton, Colorado, plant that prepared and tested the software were ``emotionally devastated'' the magazine said. Employees there were already struggling to cope with the murders of students at Columbine High School, where many of their children attended class, and the recent announcement of 900 job cuts.

The Titan 4 failure was the costliest in a string of U.S. space misfortunes. In the last nine months two satellites were blown apart in midair explosions, three marooned in the wrong orbits and another vaporized in the atmosphere. Over $3.5 billion of space hardware has been lost.

Lockheed Martin, which suffered three launch failures in April alone, and the U.S. Department of Defense last week announced separate inquiries into the rash of space mishaps.

-- Andy (2000EOD@prodigy.net), May 12, 1999

Answers

... and how do we know it wasn't software sabotage?

(bad software! bad! bad!)

-- No Spam Please (No_Spam_Please@anon_ymous.com), May 12, 1999.

Seriously, if the recent Titan and Centaur failures were software-caused, calls for programmer certification could rise if Y2k doesn't fizzle.

("Certifica-" being an optimistic alternative to "execu-".)

-- No Spam Please (No_Spam_Please@anon_ymous.com), May 12, 1999.

Now I ask you, in view of the impending, and in some cases, ongoing y2k problem, wouldn't it seem prudent to put all space launches on hold until after rollover, or until "they" can fix it. The village idiot knows this much.

I will stretch your imagination even farther, by suggesting that in view of the amount of money involved, $l.2 billion, wouldn't it seem even more prudent to suspend launhes until a thorough understanding of why the mishap occurred could be determined. The local telephone operator could tell you this is so.

And am I being farfetched to think that the launches should have been halted after the first two failures, three should have been the last nail in the space coffin, and after a fourth, fifth and sixth doesn't it seem as if they've slipped into obsessive, compulsive behavior, perhaps dotage, or a macho need to prove their spacehood. Even my dog understands this type of behavior.

And would it not be fair to shift some of the blame to the corporations who benefit from supplying the gazillions of dollars worth of parts and equipment for these launches; and to the congressmen, bought and paid for by the corporations, to promote these debacles knowing full well that y2k is a problem. A lowly downsized, file clerk gets this.

And am I being decidedly dense or queer or naive, in thinking "we the people" should stand up on our hind legs and protest at having our hard earned tax money blown away on six failed launches. Why are we screaming bloody outrage! Am I suffering from unrealistic expectations?

Thanks for the news Andy. Although I don't always agree with some of your thoughts, you are always welcome at our house for a sprout and tomato sandwich.

-- gilda (jess@listbot.com), May 12, 1999.

my immediate assumption as to why they continue to launch with supect code is that we desperately need intelligence gathering resources.

-- Cowardly Lion (cl0001@hotmail.com), May 12, 1999.

To the conspiracy buffs out there - I did notice that the "commercial" spy satellite - the one that the military was afraid would provide a "poor man's" exotic spy satellite at almost zero expense to other nations and to commercial interests - failed after launch, but they cannot apparently tell if it landed back in the ocean, in a bad orbit, or in the right orbit but with controllers "off".

If the first, the military's fears would be erased completely - by "accident". If either of the other two - what's to bet they magically "restore" communications just in time to use it for emergency military missions later, and pay off the civilian company (insurance and profit) very quietly later. Keep track of this one.

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), May 12, 1999.

Who says that they aren't there doing the work that they were supposed to do? A Military satellite not working at this time, during our non-war war. How surprised am I? Not to surprised at all. My guess is that is is working just fine.

I am not saying that mistakes do not happen, because they do. I seem to remember watching two different rockets blow up shortly after lift off because they loaded the wrong coordinates into the program. Seems like they would have come back down on the Cape. They had no options but to kaboom them.

Since the different types of rockets use different programs, I find it very unlikely that they would fail to this degree so close together. But then again, who says that it wasn't the "shrimp boat" off the coast that never shrimps? Gee... I miss working there, just might have to return.

-- (worked@there.before), May 12, 1999.

I like the Arianne rocket failure a few years back - they "reused" a control program from the Arianne 4 rocket in the faster, deeper space capable Arianne 5. But the control program for the slower rocket "ran out" of digits for a "initial conditions sensor used just at takeoff, but not during flight. (Too much accelation, or too much horizontal speed - I don't remember), When the input went to 0.00, the digital processer "reset" to 00.0000 - in both controller processers for initial launch controls and for the downwind flight controls.

The in-flight sensor, finding a 0.0 condition, interpreted this as a failure, and shut down down the main attitude controller as a safety measure - the rocket was then blown up manually when ground controllers found they could not steer the rocket. (The program error had always been present, but the 00.0000 limit was never "tested" in earlier flights because the earlier rockets were slower.)

I guess this makes the Arianne a MM failure? A 00 failure?

-- Robert A. Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), May 12, 1999.

The sabotage issue is still very much alive here because the software was corrupted by improperly loading the correct software and memory constants. Believe me I know this issue intimately, because we're going through a "military-industrial rectal exam" on our procedures for doing the final software load process for Titan components we build. The sabotage issue comes into play when asking "If all the proper steps were taken, as documentation claims to show they were, then is someone lying about what was done and why?"

What we're talking about here is either incompetance or negligence on behalf of whoever did the final software load and memory initialization, then documents being falsified to cover their mistake. Or we have someone who intentionally screwed-up the process and then falsified the documentation to try and cover their tracks to insure the error wouldn't be discovered until the bird was aloft.

WW

-- Wildweasel (vtmldm@epix.net), May 12, 1999.

CNN reported today that the next shuttle mission is on hold. Reason given was potmarks from recent hail damage.

-- john (jbhager@webtv.net), May 12, 1999.

I had to put this somewhere: it's Paco Rabanne, Paris designer, (who also sells a super-sexy aftershave) convinced that the Mir space station is going to come down and fry Paris on Aug. 11. Full of Nostradamus stuff, too.

-- Lisa (lisa@work.now), May 19, 1999.