Skip to main navigation Skip to search Skip to main content

Performance of Shrinkage Estimators for Bioburden Density Calculations in Planetary Protection Probabilistic Risk Assessment

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Planetary protection (PP) is a discipline that focuses on minimizing the biological contamination of spacecraft to ensure compliance with international policy. The National Aeronautics and Space Administration has developed a set of requirements (NPR 8715.24) based on recommendations from the Committee on Space Research that each mission must comply with regarding both forward and backward PP. Biological cleanliness requirements to target bodies, such as Mars, include spacecraft assembly control and the direct testing of the microbial bioburden of different components to comply with PP requirements. The data for each component are collected using either swabs or wipes. For each component, a number of samples are collected on one given date or on several different dates along the course of the part assembly. Given the clean spacecraft, on the average 93% of the swabs and 63% of the wipes have no colony forming units (CFU) count at 72 hours, resulting in ~85% of the 39,379 petri dishes yielding 0 CFU. Due to low CFU counts and small sampling areas, given the Poisson distributional model, the bioburden density estimates have inflated variance and confidence intervals. Shrinkage estimators are standard tools to deal with large variance and estimate inconsistencies. This paper presents the performance results of six shrinkage estimators along with the maximum likelihood, population-average, and zero estimators applied to the bioburden density estimation using InSight mission data. The results show that, for poolable data sets, the best estimator is population average, while for nonpoolable data sets, the Tsui estimator along with the empirical Bayes estimator produced the lowest mean squared error.

Original languageEnglish
Title of host publication16th International Conference on Probabilistic Safety Assessment and Management, PSAM 2022
StatePublished - 2022
Event16th International Conference on Probabilistic Safety Assessment and Management, PSAM 2022 - Honolulu, United States
Duration: Jun 26 2022Jul 1 2022

Conference

Conference16th International Conference on Probabilistic Safety Assessment and Management, PSAM 2022
Country/TerritoryUnited States
CityHonolulu
Period06/26/2207/1/22

INL Publication Number

  • INL/CON-22-66928
  • 127230

Fingerprint

Dive into the research topics of 'Performance of Shrinkage Estimators for Bioburden Density Calculations in Planetary Protection Probabilistic Risk Assessment'. Together they form a unique fingerprint.

Cite this