Abstract
Selected results of a research project are presented that investigate complexity growth during the complete development process of real-time systems. Our approach is to measure complexity factors independent of underlying development processes, specification methods, and development environments. We explain how the metrics are selected and then integrated into different modeling techniques. Metrics are collected automatically in a CASE environment and thus exemplified for real-time automation projects. Some typical complexity traces are provided and described. Based on such tracing plots it is possible to characterize projects, to compare different projects, and to control projects early depending on the evolvement of complexity.