Recently there has been some confusion about the status and processing of waivers in the Space Shuttle Program, and I would appreciate the opportunity to put some of the concerns and confusion to rest.
It has been a long established part of our process to have extensive and exceptionally well defined requirements for how we do our work... Long experience has shown that the only way to be successful in the exactly task of building, processing, and flying human space craft is to be rigorous and attentive to detail.
Whenever a detailed requirement has not been met, it is necessary to carefully consider the circumstances. Let me be perfectly clear on the first step: any time a deviation or escape from a requirement is detected it is the responsibility of each individual to bring management attention to this fact, generally through the initiation of a problem report or discrepancy report which may result in the generation of a waiver request. We must maintain this high degree of vigilance.
When a waiver request is brought to program management, I have personally adopted a three step process in reviewing any requirements escape. First, the question should always be, can we return the hardware or function to be fully in compliance with our requirements. Generally we will rework as required to achieve this standard. However, there are some cases where the rework is either not practical, not cost effective, nor could it be accomplished in a reasonable time. Experience and judgment is required here. In all cases, the Independent Technical Authority is consulted, and good engineering practices must always be observed.
If a departure from the requirements has occurred and it is impractical for whatever reason to return to the requirements, then it is imperative that good technical rationale for safety be understood. There may be tests and analysis involved. It never sufficient to list as acceptance rationale that mere inconvenience or schedule difficulties are the only, or even primary, reason for granting a waiver. No, it is mandatory that a technical measure of the adequacy of the hardware or function is sufficient to ensure safe operation at least for a limited time. In reviewing those waivers that the program had accepted before the Columbia accident, too many were granted for the life of the program with no plan to return to the requirement, and too many were granted with schedule as the only rationale for accepting an unquantified risk.
Finally, if a waiver appears to be warranted, a question must always be asked if the requirement is too stringent, if it is out of date, or otherwise needs to be revised. This is very seldom the case, but these do occasionally occur. It has been my experience that some elements interpret some requirements more stringently than other elements, and that there are requirements that are bureaucratic rather than technical, and that some requirements have been overcome by advances in technology in the 20+ year life of our program. At the program management level we are trying to eliminate this deadwood from our huge volume of requirements.
So to make sure that we are successful in space flight, I ask you to keep the process honest: follow requirements, procedures, and processes as completely as you possibly can. Whenever a circumstance causes us to fall short of the defined requirement, procedure, or process; we must always call attention to the problem and determine in a logical and consistent manner how to proceed. Most of the time, this will require rework to come back to the defined requirement or process. Sometimes, it may be acceptable to continue with a slightly deficient part or process but only after thorough community review. I ask each of you to be rigorous in this, since it constitutes our primary safeguard against catastrophe.
Now, you should know that our program has been under intense criticism from some quarters about our waiver process. Because we are much more exacting in our requirement and our waiver process than other organizations, we have more waivers. Some of our independent reviewers and indeed some of our congressional overseers believe that this means we are not as safe as we should be. I believe that our process ensures that just the opposite is true. However, there is some merit to some of their concerns and because of that; we have been reviewing the existing waivers at the time of the Columbia accident. Unfortunately, the process did not turn out to be as robust as it should have been. Many waivers cluttering the books were caused by equipment that had been redesigned, eliminated, and no longer flight worthy. Several hundred waivers retired simply because the deficient parts are no longer flown. In a couple of cases, the process was cluttered by waivers that were duplicates of other risk acceptance paperwork. In all, the total number of active, current waivers in the space shuttle program has been decreased by about 90% - from over 5,000 to less than 500. Unlike the previous large number, the smaller number of active waivers is one that program management can effectively review, periodically status, and provide resources to eliminate.
As we struggle through making our waiver process more effective, I ask you to continue to be thorough and rigorous. We can only be successful if we carefully design, test, process, and fly the space shuttle within our well established, effective requirements and configuration controlled processes and procedures. Going the second mile, I would ask you to carefully consider the requirements that we have on the books, and if improvements should and can be made, take the time to initiate the change process so that we can become even more effective in the future.
I hope this note has clarified my concern and thinking on this subject. Please feel free to contact me via email, phone, or letter at any time if you have any questions on this subject. The struggle to improve the success and safety of the space shuttle program is something that I hope you will join with me in working on every day.
Wayne Hale
Deputy Program Manager
Space Shuttle Program
