In June 1962, NASA made the decision to use a dedicated capsule to descend to the surface of the Moon from lunar orbit, rather than to descend to the surface with the Command/Service Module used to bring astronauts to lunar orbit. This decision implied that the dedicated capsule, later named the Lunar Module, would have to rendezvous in lunar orbit with their ride home and support a crew transfer between vehicles.

This decision was made in the first year of the Apollo program, seven years before the maneuver would be executed in lunar orbit. It was made before the majority of program staff was hired and before the design contracts were awarded. Yet the decision was formative; it eliminated many possible designs and gave the design teams a starting point. It guided the work of hundreds of thousands of engineers and an investment that in 1968 exceeded 4% of federal outlays.

We conceive, design, implement, and operate complex and sometimes unprecedented systems. The largest container ship today carries 18,000 containers, up from 480 containers in 1950. [1], [2] Cars built today routinely have 70 processors scattered through the vehicle, connected by as many as five separate buses running at 1 Mbit/s [3]—a far cry from early electronics buses used to communicate fuel injection at a mere 160 bit/s. Oil platforms costing $200 to 800 million [4] are developed and produced almost routinely; 39 were delivered between 2003 and 2009. [5]

These systems are not merely large and complex. They are sometimes configurable for each customer and are often very costly to deliver. Customers of consumer products expect unprecedented levels of customization and configurability. For example, BMW calculated that it offered 1.5 billion potential configurations to its customers in 2004. [6] Some complex systems are very costly to deliver. Norm Augustine points out that the unit cost of a fighter aircraft rose exponentially from 1910 through 1980, predicting that in 2053 the entire U.S. defense budget would procure exactly one aircraft. [7] Interestingly, Augustine’s prediction has held up well for 30 years: In 2010 an F-22 raptor cost $160 million, or $350 million if the development costs are included. [8]