IS MANNED MARS EXPLORATION TOO RISKY? HOW SHOULD SAFETY BE DEALT WITH?

摘要

Specific stressing factors for safety (long duration, complete isolation) exist in the Manned Mars Mission. But thereare also favourable characteristics; namely, the very presence of the crew allows repairing failed hardware, and evento reconfigure systems, while the split mission option procures redundancy without the need of supplementalhardware. Performed mission simulations illustrate these two factors.That exploration is by itself a risky proposal should be acknowledged. In fact, safety is to be properly weighted(constrained!) against program goals and constraints. To this end, complex risk and decision analysis are necessary,which combine safety with the many other project attributes (cost, cost and schedule risks, value for knowledgeacquisition…). This is not an easy task, as deciders are usually reluctant to specify valued safety goals and, over that,as combining attributes of very different natures is really tricky (e.g. is it appropriate to quantify safety value in termsof the "dollar value" of an astronaut life?).The problem appears even more intricate if we consider that, in order to achieve such a project, the rationalization ofits case should actually be aimed at two different goals: -in the short term, to get the political decision to launch andsustain the program; -in the long term, to fulfil the operational goals, while satisfying the other success criteria.In this effort, the following expertise should be applied: -agencies and industry experience gained in past spaceflightprograms; after all, we have demonstrated our capability to safely fly long duration missions; -risk analysis methodsand multi attribute decision analysis methods, which have been developed in the frame of high-stakes programs; -methods of human factors risk assessment, which are continually progressing. Many uncertainties still exist in thiscritical domain. Efforts should be developed, both through simulations and in-depth interactions between thedifferent actors.Such a comprehensive approach should allow establishing a sensible safety goal. Results of simulations seem toindicate that a level similar to that of Apollo could be realistic. Thanks to the flexibility and multiple abort options ofthe split architecture, this result could be achieved without the need of multiple redundant spaceships, thus limitingthe costs.In conclusion, a rationalized multi attribute approach of the decision problem for a Manned Mars program is the wayto best prepare for a positive programmatic decision and create the conditions of final success.