PhD defence by Allan Munck, Monday 26 June 2017 at 13:15, DTU Lyngby, building 324, room 240.
Abstract:
Embedded systems have evolved from simple mechanical devices to complex systems that include computers, radio systems, advanced software, etc. Embedded systems are increasingly becoming parts of internet-based solutions or system-of-systems. Development of such systems requires a strong focus on systems engineering in addition to the traditional specialty engineering disciplines used in hardware and software development.
Models are increasingly used in systems engineering for describing hardware, software, operators and environments of the system under consideration. To increase the quality of the models and the resulting end products, this PhD project proses a test-driven, model-based form of systems engineering.
This PhD project introduces a method for selecting modeling languages, tools and methods that are applicable to the company or organization under consideration. The method allow identifying and implementing feasible sets of tools that in combination meets all essential requirements. The method was used to select tools for a company, GN Hearing A/S.
This PhD project also introduces a method for test-driven modeling of embedded systems. The method is based on formal methods using the UPPAAL modeling tools. The method was used for modeling and analyzing essential parts of an embedded system for the company GN Hearing A/S.
This PhD project finally introduces a method for test-driven modeling of very large and complex smart systems. The method combines the use of descriptive models, formal models, informal simulation models and mathematical forecasting methods. The methods can be used for predicting the behavior of systems that are too large and complex for traditional simulation-based methods. The method was used for modeling and analyzing essential parts of a smart system for the company, GN Hearing.
The methods proposed in this PhD project facilitate structured systems engineering activities for development of embedded and smart systems in various domains.
Read more about this thesis in DTU Orbit.
Supervisor: Professor Jan Madsen, DTU Compute.
Co-supervisors:
Professor Paul Pop, DTU Compute, and Senior Vice President, R&D, Lars Linqvist, GN Resound.
Examiners:
Associate Professor Sven Karlsson, DTU Compute.
Professor Peter Gorm Larsen, Aarhus University.
ASIC Design Manager Peter Østergaard Nielsen, Sigma Designs.
Chairman:
Associate Professor Flemming Stassen, DTU Compute.
Everyone is welcome.