Figure 1. Structure of the integrated design model.
Résumé La conjoncture économique actuelle conduit les industriels à définir des produits au meilleur coût. La définition technique d'un satellite et l'estimation des coûts sont intimement liés. Dans chaque cas une base de données est nécessaire pour l'aide à la décision qu'elle soit technique ou financière. Le résultat optimum ne peut être obtenu qu'en associant étroitement les deux. Le principal obstacle étant la communication, l'ESA se devait de concevoir des outils à la fois: performants chacun dans leurs domaines respectifs parfaitement compatibles, rapides et d'utilisation facile. Au début des années 90 sont apparus des logiciels qui répondent à ces besoins et sont utilisés par les industriels Européens.
Funding:
Infrastructure Budget
The design of space systems involves integrating and optimising a number of interacting parts, each designed by specialists. In the past, this has been done by initially setting up an outline design which is then worked upon by subsystem designers. They, in turn, submit their results to the system designer who then reiterates the design, matching subsystem interfaces with customer requirements, until the design is complete. An especially critical feature is that the cost and the risks of the project are often not estimated until the design has been closed.
Recent improvements in PC-based spreadsheet and database tools have made it much simpler to create integrated system models, including cost and risk information, which capture the initial design calculations in an integrated manner. The space system being described at any time by such a model represents a closed design, and various options can be quickly compared on the basis of critical parameters like life cycle cost and overall mission risk.
ESA has recently funded a small study by Matra Marconi Space (UK) to establish an integrated design model (IDM). This is a spreadsheet-based system model, for use in conjunction with a configuration work done in- parallel with computer-aided design (CAD) tools, to support system design during early phases of a study.
ESA has also implemented a parametric database (PDB) using commercially available software, which contains all the parameters needed to make a parametric cost estimate in the early phases of a project. Having fully access to compatible tools, it becomes obvious to link them.
The objectives of the communication between databases are:
System Models have been pioneered in the USA by the Jet
Propulsion Laboratory (JPL) which now routinely uses them as
part of its integrated Project Design Centre (PDC). Within
ESA, the Working Group on System Engineering Tools amongst
others, has identified the need for a specific system
engineering tool to support the design process, particularly
during the early phases of a project. In response to that
need, the development of the IDM was initiated.
The IDM (Figure 1) provides a structured environment for
the definition of a system, as well as a standardised
knowledge base, containing known numerical parameter values
and the relationships between system parameters. Younger
design studies can inherit this data base from older designs.
The engineering information is most conveniently located in
databases, from which functional relationships between system
performance and sizing parameters can be derived. Databases
are external to the spreadsheet tool, but must be compatible
with it (i.e. ad hoc tables can be created and formatted in
the database, and exported to the spreadsheet).
Figure 1. Structure of the integrated design model.
The IDT has been based on a spreadsheet implemented in a popular commercial package. This tool is widely available and allows calculations to be laid out in an explicit, easily understandable form (with the possibility of attaching notes to particular relationships).
Because the expertise of the subsystem specialists, the risk analyst and the cost analyst is, to a certain extent, captured within the system model, this approach allows a measure of concurrent engineering to be achieved early on during a project's life time, without physical co-location of teams. It is stressed that the model cannot, and is not intended to, replace the actual engineering team and further, it needs to be complemented by additional tools (e.g. CAD). Rather, it acts as a powerful repository of the system baseline and reference at any given time and, as such, it represents the primary input to the specialists, much in the same way as large portions of a conventional system requirements document.
While the current version of the IDT is essentially a "stand- alone" system, to be used for further trials within ESTEC, context tools interacting with the system model potentially include the following:
ESA has developed a parametric database (PDB), whose contents are driven by the a parametric cost model. The PDB (Figure 2) contains historical data at unit level such as, but not limited to: mass, quantity, size, number of models, costs, and data at various levels extracted from industrial proposals and ESA's costing software system ECOS. All major proposals are submitted using ECOS, which is now used by more than 300 suppliers.
Figure 2. Structure of the parametric data base
In addition to data at unit level, PDB gives also overall information at subsystem and project level.
The PDB is menu-driven, with each menu containing several sub- menus (Figure 3). Each parameter of the PDB need be typed only once in the appropriate menu, then further selection is made using formal queries.
Figure 3. Menu Structure of the parametric data base.
Establishing interfaces between IDM and PDB is an undertaking for the months to come. The choice of working with commercial software tools, which are capable of being easily interfaced has been made while waiting for the advent of a more dedicated standard such as STEP. Due to the compatibility between commercial products, reporting is limited to a simple "copy- paste" operation. The extreme simplicity with which a query can be created provides a very effective, almost real-time, response to the IDM need for historical data.
It is essential that ESTEC acquire effective capabilities for evaluating the cost effectiveness of new missions and systems. This entails an increasingly close relationship between knowledge repositories in subsystem engineering, risk and cost analysis. Effectively pooling such resources may and should be done on the basis of powerful and convenient tools such as commercial spread sheets and databases.
Preparing for the Future Vol. 7 No. 1