Towards an Improved Understanding of Humans as the Components that Implement Systems Engineering Jakob Axelsson Volvo Car Corporation, Dept. 94100 PV 32 SE-405 31 Göteborg, SWEDEN E-mail: jaxelss5@volvocars.com Abstract. Successful systems are developed from a good understanding of two things: the features and functions desired by the customers, and the features and behaviour of the available components. A firm developing a complex technical system is also a complex technical system in itself, so the same two things need to be understood about the organisation. Much research on systems engineering focus on defining the features and functions of the organisation, but less attention has been given to the prime component used for implementing it: the human being. In this paper, an initial attempt is made at describing some of the issues involved, from a mainly philosophical perspective. INTRODUCTION Many firms produce what can be thought of as complex technical systems. These include such products as aircrafts, automobiles, power plants, and financial information systems, just to name a few. Usually, the firms are huge organisations, employing hundreds or thousands of people. To remain competitive, it is crucial for them to constantly seek to improve the effectiveness of their products. Since the products are the result of the work being done in the organisation, this amounts to improving the effectiveness of the firm itself. There is normally no difficulty in finding things in these organisations that are candidates for improvement, and there is also an abundance of general suggestions for tools, processes, and methods, that are claimed to raise the effectiveness. The academic researchers, tool vendors, and consultants who advocate these improvements usually focus on the benefits in general terms, and rarely state what the necessary preconditions are for a successful implementation. This makes it difficult for a firm to objectively assess the value of the solution in their particular case. When suggesting ways of improvement there is often a tendency to disregard the fact that product development is mainly carried out by humans, and that humans are not computers (Bollinger, 1997). This is probably the reason why many attempts to introduce ambitious prescriptive processes fail, with much disappointment to the people who designed the processes. It is also common to underestimate the cost of process changes. Again, the humans are not computers, so it is not just a question of downloading a new piece of software. The intention of this paper is to look at the implementation of systems engineering from the bottom up, by focusing on the components (i.e. humans) that actually make systems engineering happen in a firm. By doing so, we hope to gain a better understanding of the emergent behaviour of the system that develops systems, and thereby also let us improve the processes, methods, and tools used. Understanding the engineers as humans will help not only in making the processes, methods, and tools more realistic through a better respect for human limitations. It will also allow us to better utilise the strengths of people, their creativity, intuition, and ability to deal with vague, incomplete and qualitative information. Our research is still at a very early stage (and therefore somewhat speculative), so this paper will mainly describe a few of the issues we feel need to be considered, and only present some preliminary results. PRODUCT DEVELOPMENT The question that was the starting point of this research is: How can a firm producing complex technical systems improve its product development? To answer that question, it is essential to have clear definitions of the terms involved. By a firm, we mean an organisation that transforms resources into products in order to make a profit for its owners. The resources include both technical (e.g. material, machines, computers) and labor done by people. It seems reasonable to assume that people working in the firm control the transformation by making a number of decisions concerning what resources to use, and how to use them. These decisions are made prior to the actual material transformations. Thus, we define product development to be the total effort performed by the firm in order to decide and describe the characteristics of the products it will produce. Since the product development consists of a number of decisions resulting in a description, the essential activity involved is information processing. (This does not exclude e.g. the construction of physical prototypes as part of development, since the objective of those artifacts is to confirm the correctness of the information.) At any time in this process, there is a certain set of information recording the current state. This information is partly explicit, i.e. available in written documents, computer databases, etc., and partly implicit, i.e. stored in the minds of the people involved. A keyword in the initial question was improvement, so we also need to consider what the measure of effectiveness is. Since product development is an effort within the firm, and the firm operates to make a profit, the effectiveness is simply how much the product development contributes to the profit. However, profit has a credit and a debit side. The money spent is closely related to the resources used, i.e. how effectively the information is processed, and the income depends on the quality of the result, i.e. the quality of the produced information. Aspects of quality include how well the product’s properties match the needs of the customers, the ease of producing and maintaining the products, etc. COMPLEX TECHNICAL SYSTEMS The products that are discussed here are complex technical systems, and this term may need some clarification. A system is a set of interrelated and interacting elements with properties that are not exhibited by any of the elements in isolation. Those properties are sometimes referred to as emergent (Checkland 1993). It should be noted that when we purchase a product which is a system, we primarily base the decision to buy on the emergent properties of the system as a whole, rather than on what elements it consists of. However, in a buying situation we often have to rely on the information