|
Niwot Ridge Resources A Source of Information for Mission Critical Systems, Management Processes, and Strategies
|
|
Ariane 5 Failure Resources There has been much discussion of the Ariane 5 Failure and the use of XP–Like testing processes in the forums. The conjecture that test first would have saved the Ariane is pure conjecture, with little understanding of the processes by which the flight control software was and is developed. I attribute this speculation to simply not understanding the facts of the case, as well as not doing proper research on the topic before stating an opinion (this is an all too common disease of the web I'm afraid). The following resource links provide some background on the subject. The conclusion is that the requirements elicitation and management process was the root cause of the failure. It should be stated that some of the Ariane articles here have a specific "bone to pick," and a not–so–hidden agenda to sell something. By examining these resources hopefully it will become clear that the code that created the exception was:
Ariane 501 Failure Case Study
An Analysis of the Ariane
501 Failure
Design by Contract, The Lessons of Ariane The Ariane 5 Accident: A Programming Problem? Inquiry Board Traces Ariane 5 Failure to Overflow Error Ariane 5 and Design by Contract: A Reiteration of the Basic Points The Ariane 5 explosion as seen by a software engineer
ARIANE
5 Flight 501 Failure Report by the Inquiry Board
Ariane 501 Thread Archive from the 1996 safety critical systems mailing list archive hosted at the High Integrity Systems Engineering site at York University.
Here are some links to Mars Program Failures:
Previously Released: Report on Project Management in NASA by the Mars Climate Orbiter Mishap Investigation Board (Stephenson Report) - March 2000 can be found at the NASA Reports FP site ftp://ftp.hq.nasa.gov, then follow the folders to ../pub/pao/reports. This folder has all the reports for various flight missions and related data. |