Date of Award

Spring 1-1-2012

Document Type


Degree Name

Doctor of Philosophy (PhD)


Aerospace Engineering Sciences

First Advisor

David M. Klaus

Second Advisor

Alireza Doostan

Third Advisor

Louis Stodieck

Fourth Advisor

Peter Polson

Fifth Advisor

Richard Heydorn


The ability of crewmembers to perform various critical functions during spacecraft operations is widely recognized as being essential to mission success. This necessity motivates the desire to better characterize factors that can influence crewmember performance so that those with positive effects can be enhanced, while those with negative impacts can be minimized. Established Human Reliability Analysis methods exist for analyzing performance within the context of myriad terrestrial scenarios. Many of the existing methods have their roots in nuclear power plant operations. While perhaps similar, the factors influencing performance traditionally used in these methods do not take into account the unique conditions encountered during spaceflight. Therefore, this research has identified a tailored set of factors that influence human task performance during space missions. This thesis describes an organizational scheme developed to aid in classifying and communicating the factors across disciplines and organizations. Definitions of identified factors are given for the spaceflight-specific context. A visual display of the factors, called the Contributing Factor Map, is presented and its use as a risk communication tool is discussed.

The Bayesian Network is discussed as a quantification approach allowing relationships between factors, in addition to the factor relationships to performance outcomes, to be modeled. A method for determining a network structure was developed for domains such as human spaceflight, in which a global set of data for analysis is not available. This method applied the Analytic Hierarchy Process, and causal latency concepts from the Human Factors Analysis and Classification System in a novel way to guide choices for modeling the dominant set of factors and relationships in a simplified Bayesian Network structure. In addition, an approach for modeling the factors as statistical variables in a Bayesian Network making use of existing design requirements and human performance data is discussed. Applications of this modeling approach in terms of requirement completeness assessment and identification of future research needs are also described. Finally, an illustrative quantified Bayesian Network for the spaceflight domain is given, built on the factor identification and structure development work throughout the thesis. Its use in a Human Reliability Analysis is demonstrated.