An Effective Prediction Approach for Satellite Probability of Success

Reliability prediction (probability of success) using the MIL-HDBK-217 models is deemed conservative compared to actual observed reliability. This paper proposes an approach to adjust satellite probability of success prediction from the MIL-HDBK-217 models supplemented with on-orbit flight data. Using a proposed adjustment factor, this paper narrows the gap between the predicted and on-orbit demonstrated probability of success. The contribution of unit types to the adjustment factor is also developed. A Bayesian posterior estimate is used in order to avoid dramatic changes in the adjustment factor due to on-orbit data fluctuation. An example is provided to demonstrate the effectiveness of the proposed approach in predicting the probability of success. And, finally, a proposed path-forward is outlined.