Reforms and Results

Dutch physicist J. H. Bannier, former chairman of the CERN Council, headed ESRO’s external review. Bannier’s commission recommended strengthening project management by giving project managers control over technical spe­cialists and control over expenditures, ‘‘within well defined limits, in the same way as . . . the technical aspects of the work’’ were controlled. The com­mission also recommended that ESRO have more responsibility to issue and monitor contracts to relieve the Administrative and Financial Committee of trivial duties. It also advised separating policy decisions at ESRO headquar­ters from day-to-day project management and technical duties at ESTEC. Director-General Bondi defined new financial rules in November 1967 to en­sure a minimum 70% return of funding to member countries from ESRO con­tracts. By 1968, ESRO had implemented the bulk of the commission’s recom­mendations.50

Financial problems with the TD project and industry’s assertiveness spurred ESRO’s internal review, which focused on improvements to project implementation. Director-General Bondi initiated ‘‘urgent actions,’’ leading to the development of a working paper by Schalin, an ESRO headquarters ad­ministrator. Schalin’s paper, presented to the ESRO Council in March 1968, presented a blueprint for improving procedures for ‘‘forecasting, preparing for and implementing major projects in ESRO.’’ Basing his paper on ‘‘previ­ous efforts and a recent visit to NASA,’’ Schalin proposed changes to project cost estimation, contracting procedures, and project control.51

To improve cost estimation, Schalin proposed a version of NASA’s phased project planning, along with project cost-estimating formulas that GSFC ad­ministrators were developing. Schalin noted that on TD-1/2, ESRO and the MESH consortium did not develop reliable cost estimates until one year after contract award. On other contracts, a stable estimate did not occur until 75% of the funds had been committed. Schalin proposed that ESRO spend 5-10% of the total project cost for project definition and design phases to develop accurate cost estimates prior to final contract decision. He recom­mended adoption of Project Definition and Detailed Design Definition phases between feasibility studies and the development contract award. Both phases would use competitive study contracts lasting six to twelve months, using either fixed-price or cost-plus-fixed-fee contracts.

Schalin realized that past experience was the most reliable guide for early forecasts. With little experience, ESRO’s forecasts were problematic. NASA, on the other hand, had developed spacecraft cost-estimating formulas with a purported accuracy of 10-20%. ESRO administrators evaluated cost-estimat­ing techniques from GSFC and the Illinois Institute of Technology Research Institute, finding that ESRO spent a higher proportion of funding on admin­istration and less on hardware than did GSFC. This surprised ESRO admin­istrators, who had expected that their costs would be less than those of the United States. Instead, this ‘‘Atlantic factor’’ represented ESRO’s more difficult communication problems.52

Finally, Schalin recommended substantial changes to ESRO’s project con­trol methods. He noted that HEOS-A had developed extensive project con­trol procedures, which TD-1/2 implemented only after a huge underestimate. Schalin proposed a stronger contract support organization, in conjunction with rigorous change control based on detailed specifications, implemented through a network plan with work breakdown structures tied to the account­ing system. Schalin’s paper became the starting point for ESRO’s project man­agement reforms.53

By fall 1968, the Bannier committee’s organizational reforms and Schalin’s recommendations took effect. ESRO managers restructured the budget and introduced a new financial plan. ESTEC managers found that for current projects, ‘‘HEOS A2, being essentially a repeat, and the Special Project [TD-1], being further advanced, cannot be fully adapted to the new procedures.’’ For these, ESTEC increased the project manager’s authority with ‘‘extra staff con­trolling cost and contract matters.’’ It also provided more technical support

to project teams and implemented a ‘‘combined network/work package/cost control system.’’54

After considering the “possibility for the Organisation taking over the prime contractorship from industry,’’ TD-1 managers added twenty-three full-time personnel to the project, including eight for a new Project Con­trol Section, bringing the total to fifty. In addition, TD-1 used part-time staff from the technical divisions — the equivalent of forty full-time employees.55 Managers divided the MESH consortium’s tasks into 600 work packages, monitored on a monthly cycle of ‘‘data collection-processing-presentation- interpretation-decision.’’ For HEOS-A2 and ESRO-IV, which were follow-up projects to HEOS and ESRO-II, ESTEC managers used phased planning, al­though without competition.56

Phased planning crystallized into full-fledged procedures in June 1969, with the release of ESRO’s Phased Planning for Scientific Satellite Projects Guidelines. ESRO management intended phased planning to provide ‘‘clear and logi­cal build-up stepwise of information for management decisions’’ that maxi­mized project insight, yet minimized project commitment up to the deci­sion points. The guidelines defined six phases: mission studies, preliminary feasibility studies, project definition and selection, design and detailed defi­nition, development, and operations. For each phase, ESRO defined specific processes and products, which organization was to produce them, and who was to decide whether to proceed. By November 1969, ESRO’s Administra­tive and Finance Committee translated the phased planning guidelines into new procedures for satellite contracts. ESRO used the full phased planning procedure on its next satellite, COS-B.57

A second major initiative sponsored by Director-General Bondi was the development of a management information system (MIS). In August 1968, Bondi approved a proposal to establish an MIS Study Group. The study group’s efforts converged with ongoing activities at ESTEC to develop a proj­ect control system.58

Two MIS models stood out as particularly relevant: the Centre National d’Etudes Spatiales (CNES, the French space agency) chart room, and the NASA Marshall Space Flight Center (MSFC) MIS. The CNES chart room was entirely manual, where the chart room itself was the ‘‘data bank’’ of histori­cal, statistical, operational, and project information. MSFC’s system was an

Hoernke’s analogy of engineering and project control.

example of a fully computerized MIS that handled inventory, PERT project information, parts and reliability data, and videotape and library collections.59

To help determine what kind of system would be appropriate, MIS Study Group member H. Hoernke compared the concept of project control to the more general concept of control in engineering. He defined five components of an engineering control system: the process to be controlled; the sensor; the collator, which compares ‘‘what is taking place with what should be taking place’’; the memory, which records the standard for what should be hap­pening; and the effector, which changes the process toward the standard. On projects, various organizations performed the functions of the sensor, collator, memory, and effector. In his analogy, the project control group acted as the collator of data collected by various teams, and management acted as the effector.60

By November 1968, ESTEC personnel decided to use the IBM PMS 360 pro­gram for project control.61 This restricted management information to items compatible with the PMS program, including cost and schedule information stored in work breakdown structures. The IBM program could print a num­ber of standard reports, which Hoernke described in his assessment. In that same month, Hellmuth Gehriger of the ESTEC Contracts Division proposed to extend his division’s work to include “management services.’’ In early 1969, ESTEC management approved his proposal to create a Management Services Section, which would perform research on managerial problems, recommend standards, keep statistics on management performance, and aid ongoing pro­grams. Gehriger also proposed that this group, eventually called the Project Control Section, perform operations research and systems analysis studies.62

The MIS Study Group found numerous cases of redundant information generation, haphazard use, inconsistent levels of detail, and widely varying implementation of procedures and automation within ESRO. It concluded that creating an MIS would be a formidable task but provide substantial bene­fits. Aside from standard arguments that an MIS would improve management performance and organizational efficiency,63 the MIS group also argued for an MIS for political reasons. Foreseeing a possible merger with ELDO and CETS, the group deduced, ‘‘When two or more organisations merge, it is clearly the one that has the more organised structure that is at an advantage.’’ The group stated, ‘‘To control its contractors, ESRO must at least be as efficient as indus­try in managing the information problem.’’ It thought ‘‘ESRO should play a leading role in getting industry used to advanced management techniques.’’ In the opinion of group members, an MIS would give ESRO a distinct advantage in the coming bureaucratic battle.64

With Bondi’s endorsement, the MIS Study Group decided to develop a ‘‘semi-integrated system’’ where each ESRO facility would have its own com­puter system. This had the advantage of virtually ensuring the “rationalisation of information’’ at that site but the disadvantage of potentially promoting in­formation barriers between ESRO sites. A fully centralized system at Paris, the group believed, would be very complex and would politically generate ‘‘high resistance,’’ both from ESRO personnel at other facilities and from the national delegations.65

The resulting distributed computer system, called the Planning, Manage­ment, and Control (PMC) System, began operation in January 1971. At the be­ginning of a project, managers and engineers entered financial and schedule information, coded as work package numbers tied to the accounting system. The system generated reports, including internal budgets, plans, differences between plans and actual events, and changes to plans.66

At ESTEC, the Project Control Section of the Contracts Division ran the PMC System. The section’s director, Hellmuth Gehriger, became a vocal man­agerial theorist. He believed that project control was rapidly developing into a science. The project manager determined what had to be ‘‘project con­trolled,’’ and the project controllers determined how to manage the projects. Gehriger used the critical path method to schedule tasks in the phased plan­ning cycle and produced planning and control documents and information

The ESRO Planning, Management, and Control System.

flows to closely monitor project cost and schedule. From reported status in­formation, project control personnel prepared a Key Event Schedule Trend Analysis report that monitored schedule trends for slippage, a sign of im­pending technical and cost difficulties. Under Gehriger’s guidance, the Project Control Section developed a sophisticated management scheme that adapted American managerial concepts to the European context.67

ESRO management did not give its project managers complete control. In December 1969, the head of ESTEC’s Satellites and Sounding Rockets Depart­ment proposed that ESTEC assign all project personnel to his department, thereby giving him and the project manager control over all project person­nel. ESRO Director of Administration Roy Gibson refused. Although Gibson agreed that ESTEC’s use of manpower was ‘‘extravagant,’’ he did not “recom­mend the same cure.’’ Gibson believed that the proposal would ‘‘lead to the

Satellites and Sounding Rockets Department becoming an independent state within a state.’’ Instead, Gibson argued for better coordination of manpower within a matrix system. This accorded with the opinion of the head of the Personnel Department, who noted that the ‘‘home division’’ contained the ‘‘reservoir of knowledge’’ required for projects.68

Over the next four years, ESRO managers struggled with the division of authority between the project manager and the technical divisions. ESTEC established a new Programme Coordination Division to coordinate person­nel between the projects, the technical divisions, and the European Space Operations Centre. After difficult negotiations, ESTEC Director Hammar – strom issued a directive that required project managers to create and review a support plan twice each year and to request support from technical divisions through standardized forms.69

ESRO’s financial and political crises of 1967 and 1968 spurred a series of or­ganizational reforms. Under pressure from national delegations and Director – General Bondi, ESRO managers followed a path well trodden by the U. S. De­partment of Defense and NASA. ESRO adopted phased planning to provide management decision points and better cost estimates. For the TD project, it was too late to implement phased planning, so instead ESRO created the Project Control Section and implemented configuration management tech­niques, which ESRO used on all later programs. By 1971, the ESRO MIS was partially operational at ESTEC under Gehriger’s Project Control Section. ESRO personnel looked forward to the impending merger with ELDO and CETS, believing they had the organizational advantage.

The merger would soon take place, spurred by ELDO’s failure as well as the opportunities and hazards ofthe American shuttle program. On Spacelab, the Europeans’ contribution to the shuttle, the new ESA changed from being NASA’s junior apprentice to being NASA’s partner. To make this partnership work, NASA would require even further ‘‘Americanization’’ of European man­agement methods.