Practical formal verification using automated reasoning and model checking
INTAS Open Call 2005-2006 Pre-Proposal no. 4917
Started: September 2006 Ending: February 2009
Partners: RISC (leader), IeAT, Uni-Moscow, Uni-Kiev, GIC-Kiev, VIAM-Tbilisi
Official site: http://www.risc.uni-linz.ac.at/projects/intas/
Main objective: The main objective of the project is to advance the state-of-the-art in proving and checking techniques for information systems, and to apply them to concrete industrial problems. In more detail, by using practical problems for testing the different methods for proving and model checking, we will compare the effectiveness of different techniques, and we will identify the most appropriate ones, as well as the necessary adaptations, improvements, and combinations of methods which are more appropriate for solving industrial problems. In particular, we will put a special emphasis on combining automated reasoning with model checking, especially infinite-state model checking. As we want to test these methods in the context of their usage, a central task of the project is the testing and improvement of the techniques for program verification and synthesis, as well as the corresponding methods for finitary problems: circuit synthesis and verification, and reactive algorithms. An important task is the practical demonstration of the various methods in an industrial context. For this we will develop appropriate standards and tools for the integration of the methods into a coherent library, supported by an intelligent user-interface in natural language. By direct interaction with applications, new and interesting theoretical and algorithmic developments will be initiated.
Economic and impact. By using a comprehensive dissemination strategy, an important result of the project will be the increased awareness and acceptance of formal methods in industry, which in turn will have a benefic impact on the reliability of software and hardware systems, and information systems in general. The partners will be able to use the experience from the project, both theoretical and practical, for realizing more efficient tools, including possible commercial ones, and for performing formal verification tasks for industrial users.
State-of-the-art. The theoretical advances in both automated reasoning and model checking are very strong, and there are several dozens of practical implementations. However, the practical applications only begin to emerge, as there is still a widespread scepticism in the industrial environment with respect to the possibility of using formal methods. Still one sees that the more and more complex systems in use today exhibit failures which can only be corrected by such methods. The partners in the project are in the forefront of research in: automated reasoning, model checking, computer algebra, formal verification and synthesis, mathematical knowledge management, as well as in the application of the corresponding techniques to industrial problems.
Research plan. Overall, the research approach consists in improving formal methods by testing them against practical problems from industry. For this, the research will follow three main directions, each having several subtasks: 1. Proving and checking engines: clausal proving, proving in finite models, special calculi, natural style, proving for infinite-state systems. 2. Verification and synthesis: logic aspects of program verification, verification and synthesis in finite contexts, verification of specifications, model checking of infinite state systems. 3. Tools and evaluation: integration, mathematical knowledge bases, parallelization, user interface, evaluation and dissemination.