Using Tool-Supported Model Based Safety Analysis -- Progress and Experiences in SAML Development

Abstract

Software controls in technical systems are becoming more and more important and complex. Model based safety analysis can give provably correct and complete results, often in a fully automatic way. These methods can answer both logical and probabilistic questions. In common practice, the needed models must be specified in different input languages of different tools depending on the chosen verification tool for the desired aspect. This is time consuming and error-prone. To cope with this problem we developed the safety analysis modeling language (SAML). In this paper, we present a new tool to intuitively create probabilistic, non-deterministic and deterministic specifications for formal analysis. The goal is to give tool-support during modeling and thus make building a formal model less error-prone. The model is then automatically transformed into the input language of state of the art verification engines. We illustrate the approach on a case-study from nuclear power plant domain.