Technology portfolio options for NASA missions using decision trees

The portfolio allocation problem is pervasive to all R&D endeavors (i.e. all basic research sponsors need to be able to justify the anticipated incremental benefit/cost/ratios). The approach suggested is to utilize decision trees that capitalize on past experience of probabilistic fault/event tree analysis developed in the nuclear industry. A mission concept is formalized in terms of a sequence of event tree linkages and alternatives with probabilities/figures of merit of success and associated R&D costs ascribed at each link. For example, a search for life mission on Europa would involve site reconnaissance, site selection, landing, deep drilling through ice, small autonomous submersibles traversing the purported sea under ice, and in-situ life detection. Many advanced technologies not currently available would be required including long duration survivable systems (power, thermal, radiation), minimal mass autonomous systems (systems-on-a-chip, autonomous safe precision landing), life detection (including planetary protection) and communication of science data (ocean/ice/surface/orbiter/earth). For the Europa case, as an example, an event tree has been prepared in software (which means it is easily manipulated) with a variety of alternative technologies expressed. Mission objectives have been iterated with science teams; technology probabilities and costs at each link have been deduced and documented using information from Office of Space Science databases. These numbers are assumed to be the best estimates at present, which need to be reviewed and updated by NASA domain experts. The decision tree approach described, developed for an example long term (e.g. 2025) mission, is amenable to the introduction of time dependence if one is to consider investment strategies for nearer term endeavors, or programs comprised of time sequences of several projects.