Figure 1. Space systems engineering has to take into account a variety infrastructures to cope with challenging objectives and environments
The First ESA Systems-Engineering Workshop took place at ESTEC on 28-30 November 1995. Papers covering major aspects of the systems-engineering domain, and in particular software tools, were presented by participants from the space sector as well as from other industrial sectors. This article endeavours to give something of the flavour of systems engineering and the activities it embraces, as well as providing a synthesis of the ideas and information contained in the presentations and the concluding round-table discussion.
The first workshop on systems engineering, attended by over 170 participants and with over 30 papers given, was aimed at bringing together technical managers and systems engineers from the space sector, as well as from a variety of other industrial sectors and international bodies, for the purpose of exchanging knowledge, experiences and views on systems- engineering practices, techniques and informatics tools. A round- table discussion concluded the workshop with participants expressing first synthesis views on the material presented and the general scope and direction of the event.
By organising this workshop, the European Space Agency hoped to identify areas where current systems-engineering practices, in ESA and the space industry as a whole, as well as within other non-space industries, could be improved and made more efficient by the incorporation of lessons learned and state-of-the- art methodologies.
There are numerous definitions of systems engineering, and whilst they may differ in degree, they mostly emphasise that systems engineering acts across all technical disciplines to integrate and synthesise techniques in support of all phases of a project or programme.
Among the various elements of a typical space mission which must be integrated to support the mission objectives and overall mission concept, the most important are: the space segment, the launch segment, and the ground segment. Clearly then, the application of systems engineering is of interest to a large variety of programmes, all often differing widely in terms of technology, architecture, system development approach, production, nature of users service, overall life-cycle cost, and industrial organisation. An important objective is to achieve economies in carrying out traditional and new space businesses by developing efficient and affordable infrastructures.
As the accompanying panels show, systems engineering within ESA space programmes is thus a broad topic involving many disciplines, including management and technical issues as well as international and intercultural complexities. Today's trends of decreasing governmental funding, fewer projects, and smaller spacecraft make good systems-engineering practice even more difficult and important than it was in the past. In the scientific area, project teams are typically pulled in two opposite directions. Financial constraints, high reliability and schedule obligations require a relatively conservative design approach requiring low-risk strategies. On the other hand, the need to be competitive (indeed to be selected in the first place), the long periods between similar mission opportunities, as well as the desire to optimise scientific return and to remain scientifically relevant, all require an innovative, ambitious and relatively high-risk approach. Effective systems engineering must allow a balancing of these two pressures. In the end, design solutions become increasingly compromised as the margin on instrument resource budgets diminishes. In most cases, it is probably more effective to recognise the need for such compromise at an early stage so as to avoid futile work and unnecessary design features.
Now that cost is not only an issue in almost all projects, but has become a major if not the major design variable, the challenge for space system engineers is even greater than in the past. Thus, the First ESA Systems-Engineering Workshop was expected to help focus the efforts, and aid in the education of present and future space system engineers.
The Workshop showed to some extent the status of just what systems engineering is currently considered to be, where it stands in the European space sector at present, and some emerging methodologies and future trends. Although papers were presented on systems engineering in other related disciplines such as the aircraft industry, participation from outside the space sector was limited. Thus, much of what follows is specifically aimed at the space sector, although naturally the comments are still valid for other sectors.
Life-cycles and costs
The basic role of systems engineering is
to cover the global aspects of the space system, and more
generally the major mission building elements. It therefore
covers global system objectives and subsystem interfaces and is
concerned with the whole of the project, i.e. from Phase-A
(feasibility study) to Phase-E (operational) in the life-cycle.
The mission statement and purpose has changed today. The major task is no longer to make large expensive projects achievable; feasibility is not so important now as affordability and marketability. There has been a fundamental shift on objectives and cost constraints and it is now believed that more investment at the beginning of a project would allow the earlier correction of problems, which would lead to cost savings in the long run.
Systems engineering is a way to bring users, developers and operators together and in fact it must do so in order to be efficiently used in its best sense. In many applications now, it is the end-user who is becoming the driver. Clearly then, in today's systems-engineering process, more emphasis needs to be put on 'end-to-end' engineering, i.e. the process which spans from the user/market demand survey, requirements definition, design and manufacturing to delivery to the user, utilisation and eventually dismantling of the end products.
However, the end-user or customer has limited money and operates in a competitive situation. There is, therefore, a need for better optimised systems at a substantially reduced cost. There is a clear trend from requirements-driven missions to cost-driven missions, to be completed in less time than in the past. Life- cycle costs should thus be included as a direct part of each mission implementation decision throughout the project life- cycle. Much greater use of far more comprehensive commercial studies must be made to ensure that project teams have better information about the relationships of technical performance and design attributes to cost.
Figure 1. Space systems engineering has to take into account a
variety infrastructures to cope with challenging objectives and
environments
There should be a full critical review of current practices in project development and systems engineering. Where is the time spent? What is the work flow? How do people think and interact? Functions where systems engineering can help to reduce costs need to be defined. Good systems- engineering practices should enable trade-offs to be made between local alternative system approaches. Cost-effectiveness leads to competitiveness, while collaborating in order to compete is something that companies must learn to do in the present political and economic environment.
Tools and methodologies
During the workshop many papers
describing various software tools for systems-engineering
activities were presented. From the ensuing discussions, it
emerged that it was difficult to understand the precise scope of
certain of these tools, the areas where they overlapped with
other tools, and the areas for which they were specifically
applicable. It was also not clear how these tools could be used
in combination, or how much of the investment made on one tool
could be reused when moving to another one.
ESA cannot select or impose tools for industry to use. What it could do, however, is to categorise and evaluate tools and provide guidelines for their possible use and advice to the projects and to European industry on selecting the right tools. Thus, information on the appropriate tools, etc. for a given mission should be available even before a project starts. This implies developing standard databases and what is needed also are ways to measure, compare and grade the properties of systems- engineering tools. One possibility is to use faster, better, cheaper as criteria for measurement.
Tools should be seen as merely supports for systems-engineering activities, rather than the most important element. Processes such as establishing requirements, funding, interfaces, etc., are more important. It is also believed that tools do not necessarily need to be standardised and optimised the key is that the various tools should 'talk to each other' and share data through common interfaces. Data in the correct context and configuration and in the right format must be available immediately to everyone wanting to use it at the same time. To this end, the establishment of an ESA-wide Product Data Management System should be considered to provide a common interface to common and historical data.
Regarding the use of informatics tools, it can be observed that most of the current needs are covered by commercial off-the-shelf tools. However, these tools are normally closed (i.e. cannot cooperate with other tools) and therefore the use of different tools along the life-cycle implies starting from scratch because the information stored in them cannot easily be extracted for the benefit of another tool. What is needed is to make the databases these tools are using accessible so that they can be reused by other tools in terms of data and possibly even in terms of functions. Another important comment regarding informatics tools is that they ought to be designed for use by domain experts and not by software specialists.
Real-life systems engineering very often diverges much too much from the planned process, resulting in considerable cost-overruns at various levels. However, the lessons that could be learned from the past are too often simply ignored. No 'corporate memory' tool is really being put into practice so far. Common and historical data plus corporate knowledge must be available, maintained and retained and be understandable and usable by everyone involved into the mission/system design process at any level.
Tools are fairly well-developed and under-stood, but methodology is still the weak point. ESA should promote good practice and methods. This implies that they must be established first, or at least formally identified, and laid down in a set of principles. For instance, good systems engineering should minimise developments in architecture, hardware, etc. and thus modularity should be stressed. Systems engineering should also provide cost- reduction incentives. Since hardware typically accounts for only 30% of the cost of a spacecraft, there should be less emphasis on technology as the be-all-and-end- all of everything, and more importance attached to systems engineering.
Synergy with non-space sectors
Although attempts were made to
attract papers for the workshop on systems-engineering activities
from other sectors such as the automobile, aircraft, military and
nuclear industries, few were forthcoming. From the presentations
that were made, however, it was apparent that the transfer from
space to non-space sectors and vice versa has not been too good
and greater interaction is required with other sectors and the
European Union.
The military has different priorities, developments, environments and markets. So do other sectors like medical, automobile, shipbuilding, nuclear, railway, aircraft etc. The space sector should try to learn about their markets (political aspects, mass production, rapid prototyping, manufacturing, etc.), as well as their tools and innovative concepts. For example, industry asks for and uses commercial off-the-shelf equipment to save time and to be able to integrate immediately instead of first developing and making new equipment as the space sector does. This is an important lesson for the space sector to grasp.
However, there are opportunities for co-operation between the various sectors. There are several domains of potential cooperation between the defence and civilian space sectors, both of which rely on equivalent systems and related technologies in such areas as:
It is believed that synergy between the civil and military space sectors could foster the avoidance of multiple financing for the development of similar technologies and systems dedicated either to civil or military users, as well as dual-use of facilities.
In addition, space has users in the terrestrial sector (e.g. informatics, aircraft, agriculture, automobiles, etc.) where more cooperation could be fruitful. Furthermore, the design, manufacturing and production processes in the mass-market aircraft and auto-mobile industries should provide some real applications for synergy and emulation.
A number of different systems-engineering approaches and concepts were inherent in several of the papers presented during the Workshop. These included taking an integrated approach to spacecraft development, the necessity of adequate simulation and modelling, and the requirement for sound education and training for systems engineers.
Figure 2. Future space transportation will benefit from synergies
with and systems-engineering approaches from other sectors
Integrated development approaches
A number of papers were
supporting what can be seen as an 'integrated' approach to
spacecraft development, in contrast to the traditionally
'segmented' or strongly 'phase-oriented' approach. A general
tendency to want to remove the boundaries between the project
phases could be clearly identified. The motivation for this is
mainly to reduce overall schedule and costs, which can be
achieved through:
An integrated approach, while minimising overall costs, increases the costs of early phases, which are compensated by later reuse and fewer verification and integration problems.
Figure 3. Systems engineering is an end-to-end discipline
providing benefits for everyday life
The problem in implementing the advanced ideas on what could be called an 'integrated development approach' does not seem to be of a technical nature. The basic concept is defined and the tools needed to implement it are mostly available. What needs development is the methodology, which implies that the process of spacecraft development will change radically. In order to implement this methodology, project managers will have to be convinced of its worth and the right culture within industry and ESA will have to be promoted. Clearly, the present concept of geographical return poses a serious difficulty for implementing this approach.
The Prime Contractors presented a number of papers on advanced approaches to different aspects of the spacecraft development process. These approaches did not always reflect the way the companies normally carry out space projects, but rather signalled future trends. For these approaches to become established, a quantification of the benefits and savings compared with a conventional approach is needed.
The proof-of-concept should come from the application of the new approach to a real project. ESA should support the definition of such an approach and its demonstration, in order to reduce the risk for the first application, and disseminate the results obtained. Before doing this, however, the approach should be defined carefully and the necessary tools infrastructure specified and implemented with tools and methods being clearly separated.
Simulation and modelling
Rapid prototyping techniques allow for
early verification of mission concepts and subsystem interfaces.
This ensures early identification of design problems whose
correction is much more costly when discovered later. Also,
operational constraints can be taken into account at the
beginning of the project to ensure that the design will be
compliant with the mission objectives.
Establishing a simulation structure at the beginning of the project to be reused in later phases is driven by the aim of reducing the overall expenditure, i.e. the life-cycle cost, although the costs in the initial phases will increase.
Since in the early stages of new system development, extensive exploration and analysis of design options is a must, it is vital to apply a methodology for performing system design studies centring around a computerised system model containing linked descriptions of each of these components (subsystems) of the complete system. They can be further developed as knowledge about each subsystem deepens. Parametric cost-estimating models should also routinely be included as part of the complete system. Modelling, possibly supported by graphic representation, clearly can help to make system optimisation more straightforward and would also allow for subsystem trade-offs at system level.
Education and training
Emphasis must be placed on the education
and training of systems engineers to ensure that they have a
broad enough knowledge and skill base to permit them to handle
their overall task. The Workshop also showed that a lack of
mutual understanding and of commonly agreed terminology are
frequently major problems. Thus any efforts to ensure a wider
understanding and acceptance of definitions, scope, methodologies
and the like are only to be welcomed.
The knowledge and experience that systems engineers acquire during the course of their careers can also be thought of as education and on-the-job training. That knowledge will stay with them and be useful in all kinds of situations, its reuse being an essential element of the overall rationalisation principle. The reuse of knowledge is, however, an area that is lagging behind in terms of tools and practices. There is a clear need to develop knowledge repository systems and associated project procedures to ensure the proper capture and reuse of knowledge (mainly) from the early design phases in the later operational phases.
The current trend of reduced government spending on space and the emphasis on smaller spacecraft make good systems-engineering practice increasingly important. Now that cost has become the major design variable, the challenge for space system engineers is even greater than in the past. This first Systems-Engineering Workshop was an attempt to recognise and discuss these trends. Although the contributed papers focused heavily on tools for planning, designing and operating space systems, the Workshop nevertheless provided a good overview of current systems-engineering practices and highlighted the relevant aspects of the overall craft of systems engineers.
There were a number of specific proposals and suggestions made for ESA's consideration which can be translated into specific recommendations:
ESA's Technology R&D Programme could well be a suitable vehicle to trigger these activities.
Figure 4. Manned planetary bases will be the ultimate systems-
engineering challenge
The success of this first Workshop has underlined the necessity for further ongoing dialogue and exchanges of ideas in systems engineering between the various bodies and companies in the different sectors, not only space. The second Workshop, to be organised in 1997, will concentrate more on the wider issues as opposed to just tools and provide greater insight into the processes for developing and implementing cost-effective space missions.
ESA Bulletin Nr. 86.