The objective of the research project was the evaluation of new algorithms for diagnostic purposes in complex digital systems. The theoretical foundations of automated test generation integrated into dataflow networks based design systems were clarified and pilot implementations were elaborated. Experiments were carried out on the applicability of colored Petri Nets. The relation between automated test generation and integrated diagnostics was clarified. Novel algorithms were elaborated to automatically derive the starting model of this later paradigm. The foundations of the application of constraint satisfaction programming in the field of automated test generation were elaborated and applied for testing problems at the gate and functional register transfer level. The same principle was used for verification of behavioral specifications and for system-level self-diagnosis in multiprocessor systems.