There have been relatively few empirical studies of large-scale software development in Sweden. To understand the development process and be able to transfer knowledge from large-scale software development projects to academic basic education as well as to other industrial software projects, it is necessary to perform studies on-site where those systems are built.

This thesis describes the development process used at one particular Swedish company, NobelTech Systems AB, where they are trying to develop reusable software. The differences and similarities between their process and methodology and the corresponding approaches described in the literature are investigated.

The evolution of the software development process over the course of a decade is also discussed. During the 80s the company decided to make a technology shift in hardware as well as in software. They changed from minicomputer - based systems to integrated, distributed solutions, began to use Ada and the Rational Environment in the building of large complex systems, and shifted their design approach from structured to object-oriented design. Furthemore, they decided to design a system family of distributed solutions instead of specific systems to facilitare future reuse.