In the latter half of 1959, as STG monitored the gathering momentum of the various manufacturers, the urgent search for ways to reduce the ultimate risk of sending a man for a ride in an artificial moon lifted by a missile gradually became more systematic and better organized. The theme of this chapter is the quest for reliability in the automatic machinery developed for the Mercury mission. Making these devices safe enough for man took longer and exposed more doubts than STG had expected originally. During the curiously quiet first half of 1960, the flexibility of the Mercury astronaut complemented and speeded the symbiosis of man and missile, of astronaut and capsule. Technology, or hardware, and techniques, or procedures - sometimes called "software" by hardware engineers - both had to be developed. But because they were equally novel, reliability had to be built into the new tools before dexterity could be acquired in their use.1
At the beginning of 1959 NASA Headquarters had worried about three scientific unknowns needing resolution before actual attempts to conduct manned  orbital flights. In their contribution to a House Committee Staff Report prognosticating for Congress on The Next Ten Years in Space, 1959-1969, Administrator T. Keith Glennan and the chief scientists at the helm of NASA in Washington listed these imperatives that must be investigated before man could go into space:
The problems known to exist include (1) high-energy radiation, both primary and cosmic ray and the newer plasma type discovered in the IGY satellite series; (2) man's ability to withstand long periods of loneliness and strain while subjected to the strange environment of which weightlessness is the factor least evaluated; and (3) reentry into the atmosphere and safe landing. The reliability of the launching rocket must be increased before a manned capsule is used as a payload. Once these basic questions have been answered, then we can place a manned vehicle in orbit about the earth.2
By July 1959 the engineers in the Space Task Group were no longer concerned by the unknowns in each of these problematic areas. They had obviated the need for high-energy radiation shielding by selecting a circular orbit around the equatorial zone at an altitude between 80 and 120 miles, well above the stratosphere and well below the Van Allen belts. Loneliness would be no problem because the communications network would keep the astronaut in almost constant voice contact with ground crews. Weightlessness, to be sure, was the factor least evaluated, but by now this was the prime scientific variable that Project Mercury was designed to answer. The psychological outlook was good anyway, argued STG rhetorically, for does not everyone who has learned to swim enjoy the freedom and relatively "weightless" state when immersed in water? As to reentry, the strain of positive and negative acceleration forces had almost certainly been conquered; only a few questions remained unanswered about actual reentry and recovery stresses. Indeed, what Headquarters had left unnumbered in its presentation and therefore seemed to have regarded almost as an afterthought, the Task Group considered the paramount problem: the reliability of the rocket boosters must be increased before manned capsules could be attached to them.
The first major proof test of a critical part of the Mercury spacecraft design occurred on April 12, 1959. After a dismal failure a month before, the escape–tower rocket attached to a full-scale boilerplate model demonstrated its ability to lift both man and capsule away from a dangerous booster still on the ground. Giving first priority to providing an escape system in case of failure at launch was evidence of a pervading lack of confidence in the reliability of the big rockets. The men of the Space Task Group were not liquid-fuel propulsion experts; they had to rely on missile technicians and managers to convert weapon systems into launch vehicles for spacecraft. Since no one was expert in spacecraft engineering, STG had to rely on itself and on McDonnell Aircraft Corporation to gain as much experience as rapidly as possible with the capsule and its systems. This high adventure of learning how, specifically, to orbit a man safely was shared by a growing number of people supporting Project Mercury.
1 In considering how both technology and techniques began to evolve through the planning and tooling stages and into manufacturing and production, this chapter and the next make the conventional yet conceptually useful distinction between mechanical and human (factors) engineering endeavors. Another important distinction, that which rated pilot safety first and mission success second, was implicit from the start, but became explicit in the production programs only after many technical arguments and much rethinking. The process of man-rating the machines is meant to suggest all the efforts made to perfect a completely automatic system for Earth-orbital flight. The reciprocal process of machine rating men is meant to focus on the ambiguities in the idea of perfecting a completely automated system for such purposes. Chronologically this division coincides with the major, but by no means singular, concern of those responsible for the execution of Project Mercury during the year of development between the summers of 1959 and 1960.
2 House Select Committee on Astronautics and Space Exploration, 86 Cong., 1 sess. (1959), The Next Ten Years in Space, 1959-1969, report by T. Keith Glennan, Hugh L. Dryden, Abe Silverstein, John P. Hagen, and Homer E. Newell, Jr., 120.