Missle failure speculation - (Titan IV)

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

First, let me say at the beginning that I'm going to indulge in some speculation.

I doubt few Americans are aware of the recent extraordinary missle failures. In the past 18 months, there have been 3 successive Titan IV mission failures, an Athena failure and two successive failures of the new Delta III. These missions have been a mix of govt/military launches (the Titan IV) and commercial satellite launches. The latest failure (May 4) was a Titan IV-B/Centaur carrying a very large military communications satellite - the Milstar. Total cost of the mission/hardware was $1.23 billion! What caught my eye about this latest failure was an article in the May 10 edition of Aviation Week & Space Technology. The following is a quote:

"Specific data from the Milstar failure indicate the Centaur upper stage was launched carrying an inaccurate software load from Lockheed Martin that went undetected in the company's software verification process, multiple sources said."

The satellite was the most advanced military communications spacecraft ever built. It ended up in a useless orbit because of the "software" failure in the Centaur upper stage. The Lockheed software people work out of Littleton, Colorado - thats right- home to Columbine High School! The software plant had already been hurt by severe job cuts (900) and now this.

So what does this have to do with Y2K? Probably nothing - however, I can't help wondering about a few things. The US experiences a truly remarkable series of launch failures that has severely affected both our military capabilities and our commercial launch prospects. Quite a few of our enemies stand to benefit, especially China. This latest failure has the distinct smell of sabotage. The erroneous software load was discovered very quickly after the launch failure by multiple investigators. The software verification process pre-launch is VERY rigorous. Could it be something other than sabotage? Of course it could. Don't bet on it. Having said that, let me speculate on a possible Y2K link. The flight control software for the Centaur has been around for a long time, probably more than 10 years. Part of the control software has telemetry up/downlinks, including date/time data for various elements. This failure occured when the upper stage failed to follow a 3 burn profile spread over 6 hours/32 minutes. Therefore, timing information was critical and the software involved is entirely autonomous and does not take commands from the ground. It failed. The software functions off of loaded data tables (files if you like). The article said the software failed but didn't mention uploaded data. My speculation around a Y2K link would be that the "old" software had Y2K problems and was "remediated". Perhaps the "remediated" software required a new data table stucture and what got loaded was data in the "old" format. Pure speculation I know, but the string of failures has significantly hurt the US. Now how many Americans know about this? If you tok a survey on the street, how many out of 100 would know the US has taken a huge technological hit? Maybe 2 or 3 would be my guess. Just thought I would share this on the forum for your thoughts.

-- RD. ->H (drherr@erols.com), May 14, 1999

Answers

I think it has more to do with the exponential rise in software complexity and efforts to save money by reducing testing time. Lockheed had nothing to lose - the loss was completely covered by the US taxpayer.

-- a (a@a.a), May 14, 1999.

RD,

Isn't it interesting that five years ago, China was having the SAME problem with its rockets up until Clinton gave them the technology from Loral Space for campaign bucks?

Now apparently, WE cannot get any of our rockets off the ground, while China is firing 3-stage ICBM's and testing NEUTRON BOMBS WE DESIGNED BUT NEVER DEVELOPED!!!!!!

Curioser and curioser.

Someone may need to develop 200000000 Sunblock right quick, or come up with some kind of force-field shield.

Or we're all going to be glowing toast.

-- INVAR (gundark@sw.net), May 14, 1999.


Big layoffs at Boeing in Seattle, WA, and Palmdale, CA. Also Lockheed Martin in Palmdale, CA. And shooting off all "our" toys in Iraq and Yugoslavia. Meanwhile, North Korea is massing troops on the border with South Korea. Hmmm.

-- A (A@AisA.com), May 14, 1999.

My take on this situation is that the Centaur's software is loaded into its controller in the same fashion we load software into the Titan (and other) controllers we build.

Test software is loaded and run with a set of test constants (also loaded) to prove the reliability of the unit. Then the actual flight software is loaded and memory data constants are loaded for use with the test stand for the final, pre-shipment test of the unit. Once the device has completed its final test, the test constants are overwritten with data for the particular flight. Every step of the process is documented, verified and and the documentation rechecked.

At this point, it looks like the documentation was correct but the steps weren't followed. Intentional misdeed or oversight?

I imagine that now there will be a step included to have another individual perform a data download to confirm the proper data is loaded. It certainly beats having NASA flight controllers download the data and discovering that someone loaded the wrong memory constants when the controller is aboard a spacecraft that's in the wrong orbit.

WW

-- Wildweasel (vtmldm@epix.net), May 14, 1999.


To A is A@A, You said; Big layoffs at Boeing in Seattle, WA

That is normal for Boeing. They do that in cycles. Due to the fact that the unions demand big raises and promotions are almost a guarentee (especially where the peter principle comes in) The company ends up top heavy as well as the company being stuck with less than compatent workers because it is difficult to fire people. So every so often they use a lul in aircraft orders to lay off people. This time though they are doing better and time worked at the company is not protecting people, they are looking at job performance. Microsoft was more than glad to hire the doftware people who were getting laid off. In 1972 (after a big layoff due to low orders) there was a billboard that had these words on it; "Will the last person leaving Seattle please turn out the lights"

-- Cherri (sams@brigadoon.com), May 16, 1999.



Moderation questions? read the FAQ