Systems Engineering and Black Saturdays

Systems management included techniques to improve engineering and reli­ability as well as methods for managers to coordinate and control large-scale development. The formal engineering methods ultimately used by Schriever’s organization, known as systems engineering, derived largely from military programs in World War II and the early Cold War. Schriever’s group would expand upon these ideas and ensure that they were adopted throughout the aerospace industry

Although historians have yet to determine all of the originators of sys­tems engineering, many of them were involved with the military. In addi­tion, systems engineering’s proponents almost all had connections with one of two major technological universities in the United States: the California Institute of Technology or the Massachusetts Institute of Technology (MIT). At Caltech, most early systems proponents received their education under the tutelage of famed aerodynamicist and first head of the air force’s Scientific

Advisory Board, Theodore von Karman. MIT’s systems approaches stemmed from the institute’s direction of the Radiation Laboratory and other military projects during World War II.2

One the primary sources of systems engineering was the organizational culture of American Telephone and Telegraph (AT&T). Bell Telephone Labo­ratories, perhaps the single largest group of researchers in the United States outside of academia, performed research and development (R&D) for AT&T. Bell Labs researchers typically assigned hardware prototype manufacturing to Western Electric, AT&T’s manufacturing arm. Because of the large size of the corporation and the multiplicity of projects, Bell Labs and Western Elec­tric developed formal specifications and paperwork to handle the relationship between Bell Labs researchers and Western Electric engineers and manufac­turing workers. In their relationships with outside contractors and the U. S. government, Bell Labs and Western Electric personnel found it natural to use these same formal methods. In this structured arrangement coordinating re­searchers and manufacturers was the kernel of systems engineering. Donald Quarles, who headed Bell Labs for a time and later became the assistant sec­retary of defense, was familiar and comfortable with Bell Labs’ ideas about systems engineering. Mervin Kelly, who also headed Bell Labs, became an in­fluential adviser to the air force on many systems.3

MIT became involved with Bell Labs and with systems engineering in part through the Radiation Laboratory’s development of fire control systems dur­ing World War II. One major protagonist was physicist Ivan Getting, who worked on an MIT liaison committee that coordinated the integration of a Radiation Laboratory tracking radar to a Bell Labs gun director on the SCR – 584 Fire Control System. He soon realized that because of electrical noise, the two components working together behaved differently than the two com­ponents alone. Getting had to analyze the behavior of the entire system, not just its components. Because of various wartime exigencies, Getting coordi­nated the efforts of General Electric, Chrysler, and Westinghouse to manu­facture the system, acting as the de facto systems integrator and engineer for the project.4

Learning from this, in 1945 Getting made himself the liaison between the Radiation Laboratory and the navy’s Bureau of Ordnance for the navy’s Mark 56 project. He assigned the Radiation Laboratory as the system integrator for the project. The laboratory made all technical information available to Gen­eral Electric and the navy, checked and criticized designs, sent representatives to conferences, reported to the Bureau of Ordnance on progress, participated in and established procedures for prototype, preproduction, and acceptance testing, and assisted in training programs. To accomplish these functions, Get­ting arranged for the Radiation Laboratory to receive copies of navy and con­tractor correspondence, drawings, and specifications; to be notified of signifi­cant tests and conferences; to examine production designs or models; to have access to contractors and their engineers; and to inspect equipment.5 These arrangements established the formal function of system integration.6 From Getting’s position as a member of the air force’s Scientific Advisory Board and as technical director for Air Defense Command, and from weapon systems engineering courses taught at MIT, the idea of systems engineering spread throughout the air force.7

The 1949 Ridenour Report that led to the founding of Air Research and Development Command (ARDC) noted, ‘‘The role of systems engineering should be substantially strengthened, and systems projects should be attacked on a ‘task force’ basis by teams of systems and component specialists orga­nized on a semi-permanent basis.’’ Transferring authority from the compo­nent engineers at Wright Field, the report recommended that the project offi­cers and engineers who integrated components be given substantially more authority and autonomy.8 Implementing the idea took a good deal of educa­tion and exhortation, along with new regulations. Maj. General Donald Putt, a protege of Caltech’s von Karman, became commanding officer of Wright Air Development Center in 1952. He admonished the laboratory chiefs, ‘‘Some­body has to be captain of the team, and decide what must be compromised and why. And that responsibility we have placed on the project offices.’’9 Engi­neering personnel in the project office acted as systems engineers, with the responsibility for the integration of technologies into the weapon system, whether aircraft or missiles.

Systems engineering also played a prominent role at Hughes Aircraft Com­pany, where Simon Ramo had assembled a skilled team of scientists and engi­neers to develop electronic gear for military aircraft and the innovative Fal­con guided missile. The Falcon differed from contemporary air-to-air missiles in that it used sophisticated electronics to guide the missile to its target and hit it. Other missiles typically placed a large warhead near an enemy aircraft, then detonated it nearby using proximity fuzes. These required substantial amounts of explosives and hence also a big, heavy missile to carry them. Ramo and Dean Wooldridge instead used what they called the systems approach to determine a more optimal design for air-to-air combat.10

Like MIT’s Getting, Ramo formulated his notions of systems engineer­ing through work on complex military projects. Although he had worked at General Electric, where a number of organizations had the word ‘‘system’’ in them, his work on various components did not stimulate any interest in the processes of engineering. Moving to Hughes Aircraft, and soon heading his own organization devoted to military electronics and missiles, Ramo began to think more seriously about the processes common to Hughes’s varied tasks. Wondering how best to formulate and pass on the expertise necessary to ad­dress the complexities of missiles and electronic systems, Ramo began to pro­mote the idea of an academic discipline of systems engineering. However, his first opportunities to pass along these ideas came not through publication but through his involvement with Schriever’s ICBM program.11

Ramo’s company came into being as a result of a meeting between Ramo and Secretary of Defense Charles Wilson in 1953. At that meeting, Wilson ex­pressed displeasure that the eccentric Howard Hughes had captured a near monopoly on aircraft and missile electronics through Ramo’s group. Wilson informed Ramo that he intended to ‘‘break this monopoly’’ and would sup­port Ramo if he separated from Hughes. This catalyzed Ramo and Wool­dridge’s decision to form their own company, a decision soon rewarded when Deputy Secretary of the Air Force Trevor Gardner awarded them a contract to support John von Neumann’s ‘‘Teapot Committee’’ (chapter 2). Gardner was an old friend of Ramo’s, but despite this, Ramo and Wooldridge did not really want the ICBM systems engineering job because they correctly perceived that this would hinder their efforts to land lucrative hardware contracts. When the air force informed them in early 1954 that they would not acquire any air force contracts unless they took the ICBM systems engineering job, R-W accepted the contract from Schriever’s group.12

Schriever fostered close working relationships between R-W, the WDD of ARDC, and the SAPO of Air Materiel Command (AMC). Schriever and Ramo agreed that R-W personnel should be placed in offices adjacent to those of

Подпись: Image not available.

Brigadier General Bernard Schriever and Dr. Simon Ramo at a building dedication at the Inglewood complex in 1956. Courtesy John Lonnquest.

their WDD counterparts. For example, the office of Schriever’s technical di­rector, Charles Terhune, was next to that of the R-W technical director, Louis Dunn. At the highest level, Schriever and Ramo were in frequent contact.13

Despite the close contact, the function of R-W personnel was not clear to Schriever’s group as late as April 1955. Schriever directed Ramo to assemble a briefing to describe for his officers and contractors the processes and tasks that R-W performed.14 This briefing was one of the earliest descriptions of systems engineering. R-W formed its Guided Missile Research Division (GMRD) in 1954 to handle the technical aspects of the ICBM programs. With Ramo head­ing the division and Louis Dunn, the former Jet Propulsion Laboratory (JPL) director, as technical deputy, the GMRD in April 1955 had five departments: Aeronautics R&D, Electronics R&D, Systems Engineering, Flight Test, and Project Control. While the Aeronautics and Electronics departments concen­trated on subsystems and components, the Systems Engineering, Flight Test,
and Project Control departments performed the bulk of ICBM integration tasks.15

Technical direction of contractors took place through monthly formal meetings as well as numerous informal meetings. R-W Project Control per­sonnel chaired the formal meetings, set the agenda, recorded minutes, and presented current schedules and decisions. Based on the results of these meet­ings, the Project Control Department issued Technical Directives, work state­ments, and contract changes. WDD officers reviewed Technical Directives, along with changes to work statements. They then submitted work statements and contract changes to the SAPO, whose officers then issued contractual changes and approved work statement modifications. Informal meetings were for “information only,’’ and WDD and R-W personnel coordinated this infor­mation as necessary. The Project Control Department handled official plans, schedules, work statements, cost estimates, and contract changes.16

Engineers in R-W’s Systems Engineering Department analyzed major de­sign interactions, studied electrical and structural compatibility between sub­systems and contractors, and issued top-level requirements. One good ex­ample was the nose cone trade study that cut Atlas’s mass in half. Another was an assessment of the Martin Company’s trajectory analysis. Department members found that Martin’s trajectory was less than optimal; by modify­ing it, R-W engineers increased the Titan’s operational range by 600 miles, the equivalent of saving 10% of its mass. R-W systems engineers performed experimental work in the laboratory when they needed more information, analyzed intelligence data on Soviet tests, and programmed early missiles. As noted by one critic of R-W, the engineers often double-checked contractors to avoid ‘‘errors, mistakes, and failures.’’17

By October 1956, the WDD and R-W came to a legal agreement about what systems engineering entailed. The agreement defined systems engineering in terms of three functions:

1. The solution of interface problems among all weapon system subsystems to insure technical and schedule compatibility of the systems as a whole.

2. The surveillance over detailed subsystem and over-all weapon design to meet Air Force required objectives.

3. The establishment and revision of program milestones and schedules, and

monitoring of contractor progress in maintaining schedules, consistent with sound technical judgment and rapid advancement of the state of the art.18

From 1953 through 1957, R-W’s role grew dramatically. Starting with docu­mentation of the Teapot Committee’s deliberations, R-W acquired a contract with Schriever’s new organization to perform long-range studies of ICBMs, to assess new technologies, and to help the WDD set up and operate its new facilities. Its funding grew from $25,494 through June 1954, to $833,608 from July 1954 through June 1955, and to $10,095,545 from July 1955 through June 1956. As R-W’s competence grew, Schriever expanded its role. R-W double­checked contractors’ work; controlled specifications, schedules, and other paperwork; and surveyed the technical horizon for new technological solu­tions. As Schriever himself later admitted, R-W became for the WDD what Wright Field and its component engineers were for aircraft development. For the first few years of expansion, R-W’s services were indispensable to the WDD; they cut program costs and improved ballistic missile performance.19

Along with systems engineering, Schriever initiated other methods to man­age the program. As he well knew, the system approach required planning for the entire weapon life cycle from the start of the program. One of Schriever’s first actions was to establish a centralized planning and control facility to facilitate application of this idea. The WDD established its own local and long­distance telephone services, including encrypted links for classified informa­tion and teletype facilities.

In the fall of 1954 Schriever and his staff developed a management control system. Every month, they required the air force, R-W, and associate contrac­tors to fill out standardized status report forms. One of Schriever’s officers controlled and updated the master schedules, placed on the walls of a guarded program control room. This room was both a place where managers could quickly assess the ‘‘official’’ status of the program and a place where Schriever and his deputies showed the program status and innovative management to visitors.20

A primary benefit of the management control system was the process of preparing the weekly and monthly status reports. Report preparation required that managers collect and verify data, identify problems, and make recom­mendations about how to resolve them. Schriever instituted monthly ‘‘Black Saturdays’’ for project officers to report difficulties. At these meetings, Schrie­ver and his top R-W and military staff reviewed the entire program and as­signed responsibility for resolving all problems to individuals there. These meetings endeavored to bring problems forward instead of sweeping them under the rug. As Schriever put it, ‘‘The successes and failures of all the de­partments get a good airing.’’21

While Black Saturdays brought some order to the technical aspects of the program, the Procurement Staff Division of the Ballistic Missiles Office at air force headquarters had to cope with the legal and financial mess created by Schriever’s disregard for standard processes. The financial officers insisted that ‘‘the technical directives [be] covered by cost estimates’’ because annual fund­ing from the DOD was insufficient to cover rising costs. Schriever fought these regulations as ‘‘examples of the ‘law’s delay’ ’’ but had to give in. In November 1956 he agreed to submit cost estimates, leading to new procedures in Febru­ary 1957. To ensure that R-W and the other contractors documented technical directives, the Guidance Branch of the WDD in October 1956 ‘‘began holding a contract administration meeting immediately after each technical directive meeting.’’ By January 1957 the Procurement Staff Division extended the prac­tice to all technical direction meetings.22

With these new procedures to coordinate the legal and financial aspects of ICBMs, the air force could map out the ramifications of the various changes to the ICBM programs. Although this allowed for a modicum of order across the air force, only upcoming missile tests could determine whether the Atlas and the Titan would fly.