TOCS Event-Silicon Valley Campus - Carnegie Mellon University

TOCS Event


Bernard Steffen

Bernhard Steffen
Chair of Programming Systems and Compiler Construction, University of Dortmund


September 4, 1:30 pm



CMUSV, Rm 118 [directions]

[login as guest]

Title: Constraint-based variability modeling
Abstract: Mastering variability is central for reducing total cost of ownership of systems, which, due to their popularity require continuous update or specific customization. Thus topics like variability modeling and product line engineering emerged, typically organized at the architectural level in a bottom-up fashion. In contrast, constraint-based variability modeling takes a top-down approach: based on the set of ontologically characterized, reusable components, behavioral constraints define how the components may be combined to form valid system variants. These combinations can be automatically determined using an automated synthesis algorithm. In particular, combined with feature-oriented problem space variability modeling, this approach subsumes feature-oriented software development of product lines. This will be illustrated along a comprehensive case study in the context of bioinformatics.
Speaker Bio: Bernhard Steffen became 1993 Full Professor for Programming Systems at the University of Passau. He holds the Chair of Programming Systems and Compiler Construction at the University of Dortmund. He is author of over 200 refereed papers concerning various aspects of formal (verification) methods and tools for program analysis, compiler optimization, model generation, testing, and service-oriented software development. He is founder and Editor in Chief of Software Tools for Technology Transfer (STTT), Springer Verlag, and co-founder and Steering Committee Member of TACAS, the Int. Conference on Tools and Algorithms for the Construction and Analysis of Systems. In 2004 he co-founded ISoLA (Int. Symposium on Leveraging Applications of Formal Methods, Verification and Validation). In 1989 he co-developed one of the earliest formal analysis tools for distributed and parallel systems, 1991 his paper Data Flow Analysis as Model Checking helpted start the field of Software Model Checking, 1992 he presented the first functioning Model Checker for infinite-state systems with Model Checking for Context-Free Processes, and in 2002 he obtained the Most Influential PLDI Paper Award for Lazy Code Motion. His eXtreme Model-Driven Development (XMDD) paradigm, is aimed at continuously involving the customer/application expert through a systems' life cycle where model based testing is not practical. XMDD combines service and aspect orientation, model driven design and ideas from eXtreme programming in order to achieve agility. Continuous quality assurance is guaranteed via model-checking-based verification and the combination of model-based testing and automata learning.