From Joseph Fragola, Science Applications, International Corporation, New York
Your article on the space shuttle indicates that risk analysis results predict a 50 per cent probability of another shuttle loss by 2015, if shuttle usage remains constant (This Week 2 September). In reality, 51 per cent of the main engine launch risk is embodied in main engine features whose design changes were not included in the current estimate. If these and other design changes address the risk, then shuttle performance would be expected to better the current estimate.
Secondly, both the initial 1 in 78 and the 1 in 248 odds cited are median estimates of launch risk only and are therefore comparable. While the current study included total mission risk, its median estimate was 1 in 145 missions.
Finally, at the end of the article our report was misquoted; the following critical words in italics were omitted. The actual quote referred to the automated data screening process we had preferred to use for our study. This process did not prove practicable. “This was because record codes which could have more effectively screened the records were missing, incorrect or inconsistent in many cases.”
We did not find the data itself lacking. In fact, the shuttle engine data set was quite comprehensive and complete compared to other data sets SAIC has reviewed.
Advertisement
