Over the last decade, software has become an integral part of many products with which it is not traditionally associated (e.g., automobiles, medical equipment, home appliances, etc.). This has led to problems in many organizations, since it has proved difficult to integrate software engineering processes with other engineering processes. To address the increased complexity and to coordinate their engineering efforts, many organizations working in such domains have therefore introduced systems engineering into their business processes. Systems engineering is an interdisciplinary approach to system development. Teamwork is essential in developing complex systems, and systems engineering orchestrates this process throughout the lifespan of a system. The focus of systems engineering is on defining customer needs and required functionality, documenting requirements, synthesizing a design and validating the system. A problem with the traditional techniques of systems engineering is, however, that they provide inadequate support for achieving high levels of reuse between different projects. In this dissertation, methods and tools with the overall purpose of providing an improved reuse infrastructure for systems engineering artifacts and their resulting detailed design artifacts are proposed. The proposed methods and tools are based on the software product-line approach, a reuse strategy which focuses on families of related systems that address particular market segments. Studies have shown that if an empirical study does not report experiences from that which practitioners consider a “real” situation, results are likely to be ignored. Even though case studies cannot achieve the scientific rigor of formal experiments, case studies can provide enough information to decide if a specific technology will benefit a particular organization. To ensure industry-relevant research results, case studies on real development projects were chosen as the main vehicle for performing research. This dissertation reports experiences from four empirical studies which were performed within the Swedish defense industry. The studies indicate that the proposed methods and tools indeed provide an improved reuse infrastructure, compared to the previously used methods and tools.

The goal of this research is to develop a new paradigm integrating the practices of business process modeling and use case modeling. These two modeling approaches describe the behavior of organizations and systems, and their interactions, but rest on different paradigms and serve different needs. The base of knowledge and information required for each approach is largely common, however, so an integrated approach has advantages in efficiency, consistency and completeness of the overall behavioral model. Both modeling methods are familiar and widely used. Business process modeling is often employed as a precursor to the development of a system to be used in a business organization. Business process modeling teams and stakeholders may spend months or years developing detailed business process models, expecting that these models will provide a useful base of information for system designers. Unfortunately, as the business process model is analyzed by the system designers, it is found that information needed to specify the functionality of the system does not exist in the business process model. System designers may then employ use case modeling to specify the needed system functionality, again spending significant time with stakeholders to gather the needed input. Stakeholders find this two-pass process redundant and wasteful of time and money since the input they provide to both modeling teams is largely identical, with each team capturing only the aspects relevant to their form of modeling. Developing a new paradigm and modeling approach that achieves the objectives of both business process modeling and use case modeling in an integrated form, in one analysis pass, results in time savings, increased accuracy and improved communication among all participants in the systems development process. Analysis of several case studies will show that inefficiency, wasted time and overuse of stakeholder resource time results from the separate application of business process modeling and use case modeling. A review of existing literature on the subject shows that while the problem of modeling both business process and use case information in a coordinated fashion has been recognized before, there are few if any approaches that have been proposed to reconcile and integrate the two methods. Based on both literature review and good modeling practices, a list of goals for the new paradigm and modeling approach forms the basis for the paradigm to be created. A grounded theory study is then conducted to analyze existing modeling approaches for both business processes and use cases and to provide an underlying theory on which to base the new paradigm. The two main innovations developed for the new paradigm are the usage process and the timebox. Usage processes allow system usages (use cases) to be identified as the business process model is developed, and the two to be shown in a combined process flow. Timeboxes allow processes to be positioned in time-relation to each other without the need to combine processes into higher level processes using causal relations that may not exist. The combination of usage processes and timeboxes allows any level of complex behavior to be modeled in one pass, without the redundancy and waste of separate business process and use case modeling work. Several pilot projects are conducted to test the new modeling paradigm in differing modeling situations with participants and subject matter experts asked to compare the traditional models with the new paradigm formulations.