When determining criteria for decision matrix software for NASA one must consider the over all project needs. If Artificial Intelligent computers are to help make important decisions that could mean life or death for lunar colonies, astronauts or base stations, they must be honest, programmed using reality based criteria and be void of human emotion.

If we are design a business decision matrix for the best type of system to provide life support for the colony, thru mining local resources to make water, oxygen and hydrogen and other gases for fuel to create a self sustaining system then these criteria must take into account many critical factors such as;

1.) Weight of the machines needed to be transported for Mining

2.) Cost of components

3.) Shape of components due to shipping

4.) Number of robots needed to build facility

5.) Power supply needed to run system

6.) Waste disposal of anything not used

7.) By-product uses and storage

8.) Time to Build Components, Robots and System

9.) Simplicity due to Murphy'ism

10.) Miscellaneous hazards and dangers

11.) Deployment timeline

These of course will not be the only important criteria, there will be others; what else needs to be considered; which criteria would you give the most consideration? If we are to set lets say a numerical number to each criteria for our decision matrix how much should each get and why? If NASA reviews 1000 possible proposals then you can see why such an artificially intelligent decision matrix system is needed and why it is paramount to keep the system fair so that not necessarily does the lowest bidder win the right to design and build the system, but that the structural integrity is sound and can be free from Murphyism. Think on this.