Setting the Bar for the Replacement of the Probability of Collision Metric in Conjunction Assessment

Professional statisticians can often see issues with, and propose better solutions for, statistical methods that are used in industry; but such proposals often remain unimplemented (and therefore unused) because they are not fully sensitive to the particulars of the application.  The Probability of Collision, a metric used by astrodynamicists to determine whether two satellite are likely to collide, has a number of problematic aspects that statisticians have properly criticized; but the proposed alternatives also introduce fresh problems or ignore other aspects of the operational situation.  The present colloquium introduces this statistical problem and summarizes its shortcomings and some of the proposed remediations, but then proceeds to outline philosophical aspects of the operational situation that both exclude certain approaches and push towards particular ways in which proposed alternatives must be evaluated in order to be shown to be both operationally superior and viable.

Dr. Matt Hejduk is Senior Project Leader at the Aerospace Corporation and also presently serves as chief engineer for the Satellite Conjunction Analysis and Risk Assessment (CARA) project for NASA Goddard Space Flight Center.  His research interests include the satellite conjunction assessment problem, satellite characterization and modeling from sensor signature data, estimation theory, and sensor and network simulation.