There is a range of possible approaches to structuring and managing a project of the magnitude of GSMT in terms of the extent to which the central project team does the work in-house, or delegates the work to outside contractors. At one extreme, it is possible to turn the entire project over to a prime contractor. A successful example of this type of organization is the Subaru Project.
The main advantage of using a prime contractor is the option of choosing a company with professional managers and a well-trained engineering staff experienced in running large projects on schedule and on budget. Although the contractor will certainly add profit into the cost of the project, this can potentially be offset by cost savings resulting from administrative efficiency. Using a prime contractor can work well if the technical requirements and operating constraints (such as site characteristics) are well understood and can be expressed clearly at the start, so the contract can be bid as a fixed price.
The difficulty for a project like GSMT is that the technical specifications and, to some extent, the science goals, will evolve during the design and development phase, particularly as alternative configurations are evaluated during design-to-cost exercises. It is best if these tradeoffs can be made in close collaboration with the scientists and without requiring amendments to the contract.
Most large telescope projects in recent years have adopted an approach in which a central team of scientists, engineers, and managers has organized the work and developed the main design concepts, then relied on contractors and partner institutions to develop detailed designs and fabricate and deliver key subsystems. The central team then leads the system integration and testing work.
Because it is likely that GSMT will be a partnership involving multiple institutions and possibly multiple countries, this draft management plan is modeled after the Gemini Project. This project involved seven partner countries and successfully built two 8-m observatories with relatively complex associated facilities. Gemini had a medium-sized central project team (smaller than the European Southern Observatory VLT (Very Large Telescope) team, for example, but larger than the team that built the Magellan telescopes). The size of these teams reflects fairly well the relative complexity of the complete facilities that were built, but was also determined somewhat by the subcontracting approach that was followed in each case.
The Gemini Project staff was organized into a number of technical groups. Although some groups shifted over the years as the project organization was adjusted to fit changing circumstances, the following list defines the main groups:
- Systems engineering
- Building and enclosure
Each group had a group manager, who was also expected to be a technical contributor, along with engineers, designers, and technicians as appropriate.
The staff of the Gemini Project included ~ 35 people during the design phase. This number ramped up slightly during the manufacturing/construction phase. Additional personnel responsible for specific work packages were involved in the scientific organizations in the partner countries.
Although GSMT will be a significantly larger project than Gemini, the management model for this draft plan envisions a central staff similar in depth to the Gemini design staff, but with three exceptions. First, GSMT will require a site-testing group (Gemini did not). Second, the greater control complexity is reflected in having both a controls group and a software group. Third, the high reliance on adaptive optics (AO) is reflected in having a separate AO group. The draft staffing plan is described under section 6.2.5 below.
This plan assumes that the central team will be co-located during the design and development phase, because this is normally the arrangement that maximizes efficiency and minimizes risk. It also assumes that the project manager and project scientist will be recruited first, and they will in turn supervise the hiring of the project staff.
The scope of the design and development (D&D) phase includes all the activities required to get to the point of being ready to issue the major fabrication and construction contracts. This includes:
- Organization of the project
- Staff recruitment
- Technology development studies
- Site selection
- Conceptual design
- Preliminary design
- Detailed design
- Vendor selection
- Preparation of proposal for construction funding
- Preparation of acceptance test plans and integration plans
- Initiation of contracts for preproduction activities (for critical path contracts)
These tasks are described in greater detail further in this section.
The following first-level work breakdown structure starts at the beginning of the partnership, assuming the partner institutions have agreed on the following by that time:
- The nature of the partnership agreement
- The GSMT scientific context and principal scientific goals
A brief description of each task is given below, followed in the next section by a draft schedule that shows the sequencing of activities.
- Prepare and sign partnership agreement
- Recruit director, project scientist, and project
The director, project scientist, and project manager must be selected in a manner acceptable to the partner institutions. This plan assumes that an acting director is appointed at the start of the project to handle these recruiting and other start-up activities. The selection process will be conducted by search committees, including representatives from each partner.
- Establish project office
The site for the project office must be agreed upon by the partners, and a suitable facility provided. It should be set up with appropriate infrastructure, including communications capabilities and computer systems. Initial legwork can be started before the partnership agreement is signed, then continued under the direction of the project manager once he or she is recruited.
- Set up project organization
This involves a series of planning and organizing activities as listed below. These will be revised, and enhanced versions of earlier preliminary plans will be developed before the partnership is formalized. The project plan should be approved by the board of directors.
4.1 Prepare work breakdown structure (WBS)
4.2 Prepare project schedule, including task dependencies and task durations
4.3 Prepare resource plan (primarily a staffing plan)
4.4. Prepare cost estimate (for D&D project, not construction)
4.5. Prepare design-to-cost plan
4.6. Set up project budget
4.7. Prepare project plan based on WBS, schedule, resource plan, cost estimate, design-to-cost plan, and budget. The project plan will also include:
4.7.1. Organizational plan
4.7.2. Quality plan
4.7.3. Communications plan
4.7.4. Roles of advisory committees
- Recruit project staff
The project manager will have overall responsibility for recruiting, with guidance from the project scientist. Group managers will be recruited first, and they will in turn recruit the technical staff. A goal will be to recruit a number of highly qualified technical personnel from each partner institution, augmented by well-qualified personnel from industry and academia.
- Set up management systems
These systems will include:
6.1. Risk management system
6.2. Change control system
6.3. Schedule control system
6.4. Cost control system
6.5. Quality control system
6.6. Performance reporting system
- Set up systems engineering structure
This will include:
7.1. Scientific merit function
7.2. Error budget structure
7.3. Design-to-cost structure
7.4. Interface control documents
- Develop science requirements document
A formal science requirements document should be prepared and agreed upon by the partners, based on the agreed science context and scientific goals. This document should also define the scientific merit function (see Section 5.7).
- Prepare flowdown of science requirements
Once the system architecture has been defined, error budgets traceable to the science requirements will be prepared. An initial design requirements document will also be prepared for each major subsystem. These will be iterated as the conceptual design is developed.
- Establish site selection criteria
An agreement on site selection criteria, acceptable to all partners, must be reached at the start of the project.
- Test sites
This effort may require further development of site testing techniques and construction of site testing equipment. It is assumed that data will be recorded at more than one candidate site.
- Develop conceptual design
During the first phase of the conceptual design, we will consider alternative designs in order to define the system architecture, and answer such questions as: Will the facility be a single telescope or an array? Will the design be fixed elevation or fully steerable? Will it have a filled aperture? The most promising designs will be optimized to yield the greatest scientific value at minimum cost, and a preferred design will be chosen.
- Conduct conceptual design review
This meeting will review the most promising design alternatives and confirm the selection of the proposed design concept. It will ensure that this concept is consistent with the project scope, science requirements, and project budget. The review will also evaluate the systems engineering structure, the design requirements documents, and the interface control documents.
- Reduce site testing data
This is reduction of the site testing data to provide the data products necessary to make the site selection.
- Select site
The observatory site will be chosen based on the site testing data and other appropriate considerations, such as relative cost.
- Initiate preliminary design contracts
Much of the preliminary design work will be performed by subcontractors or partner organizations (for example, the design of the science instruments). This task includes the preparation of bid documents, selection of subcontractors, and negotiation of contracts.
- Complete preliminary design work
This task includes design work done by the central project team as well as subcontractors. Each subcontractor or responsible engineer will have both performance and design-to-cost requirements to meet.
- Initiate technology development contracts
This task includes the preparation of bid documents, selection of subcontractors, and negotiation of contracts.
- Conduct technology development contracts
These contracts will support the development of fundamental technology needed for GSMT, which could include, for example, development of AO technology (lasers, deformable mirrors, etc.) or development of methods for low-cost fabrication of off-axis aspheric mirrors.
- Develop subscale prototype(s)
Because of the complexity of control systems required for GSMT, some prototype testing is advisable to minimize the time required to debug software after construction is completed. To avoid this delay and the associated high costs, one or more subscale prototype facilities should be built to test hardware and particularly software. The prototypes could be one or more laboratory test beds. It may even be possible to reach an agreement with another observatory to use a smaller telescope for testing some control systems, as the NTT (New Technology Telescope) did for the VLT.
- Initiate preliminary site work
This includes applying for the necessary permits to begin the site work, and selecting qualified firms to do geotechnical surveys and prepare site development plans.
- Perform preliminary site work
This task includes geotechnical surveys and developing plans for the site development.
- Conduct preliminary design reviews
Preliminary design reviews will be held for all subsystems of the observatory. This will include review of design work done by subcontractors as well as by project staff.
- Initiate detailed design contracts
This task includes the preparation of bid documents, selection of subcontractors, and negotiation of contracts.
- Perform detailed design work
This task includes some design work done by the central project team, but most of the work will be done by subcontractors.
- Prepare preliminary proposal for construction
To avoid unacceptable delays in the overall project, potential funding agencies should be presented with a preliminary version of the construction proposal in advance to allow adequate time for review.
- Critical design reviews
Critical design reviews will be held for all subsystems of the observatory.
- Select principal fabrication and construction
Because the vendor selection for the principal fabrication and construction contracts is likely to be a lengthy process, and because the construction cost budget will be strongly dependent on the costs negotiated for these procurements, the vendor selection should be completed before the proposal for construction funding is submitted. This task includes preparation of bid documents, selection of subcontractors, and negotiation of satisfactory contract terms. Vendors will be asked to guarantee their bids for an extended length of time (one to two years) in case there is a delay in construction approval.
- Prepare full proposal for construction funding
Based on the results of the critical design reviews and the best bids and estimates available for the fabrication and construction work, the final version of the proposal will be submitted to funding agencies.
- Prepare acceptance test plans
After the critical design review, while waiting for construction funding approval, project staff will prepare acceptance test plans for all subsystems whose production will be directly supervised by project staff. For subsystems produced under contract, the vendors will be required to prepare acceptance test plans.
- Prepare integration plans
During this period, integration plans will be prepared for all subsystems that are part of the observatory facility.
- Initiate pacing contracts for preproduction
It will likely be necessary to initiate contracts for the first preproduction phases of the pacing contracts before full construction funding has been approved, with options to invoke the full contracts after funding approval. These initial contracts will enable vendors to start setting up production facilities, for example.
The partnership agreement will be a contract that clearly delineates the nature of the partnership, the participation of partner institutions, the scope of the project, the management structure (i.e., the membership of and rules governing the board of directors), and the project deliverables.
Table 1 provides a representative list of personnel for the central project staff in this management plan. This staffing level is based on the schedule shown above, with the work proceeding from project start to complete readiness for the fabrication and construction phase in five years. This staffing level assumes that the project team is responsible for performing and supervising the conceptual design and development work for the complete point design telescope, including AO systems, instruments, and necessary supporting equipment and facilities. In general, the detail design and fabrication of key subsystems will be contracted to private sector and university contractors.
The initial cost estimate for the design and development phase is summarized in Table 2. The estimate is based on the scope, staffing plan, and schedule described above, including the assumption that all key project personnel will be co-located. We believe this schedule is close to the optimum pace to minimize the total cost of the project. If the schedule is delayed because of cash flow considerations, for example, the total cost will likely be higher.
Costs are estimated in year 2001 US dollars. Inflation of salaries has been estimated at 3% per year (the US consumer price index has averaged a 2.7% yearly increase over the past ten years). Travel costs are estimated at 10% of salary costs. Other expenses, a category that includes supplies and materials, purchased services, and noncapital equipment, has been estimated at 20% of salaries. General and administrative rates are based on current NOAO rates and include 9% applied to the total of salaries, benefits, travel, expenses, contracts, and major purchases, plus an additional 11% facilities charge on salaries. However, using NOAO rates as a basis for costing does NOT indicate a presumption that the project office will reside at NOAO. Rather, its location should be decided jointly by partnering institutions.
Table 1 Staffing plan for the design and development phase. A yellow square indicates that the person is employed that indicated quarter. Alternating squares for administrative assistants indicate shared personnel between two groups.