Revisiting the Steam-Boiler Case Study with LUTESS : Modeling for Automatic Test Generation

LUTESS is a testing tool for synchronous software making possible to automatically build test data generators. The latter rely on a formal model of the program environment composed of a set of invariant properties, supposed to hold for every software execution. Additional assumptions can be used to guide the test data generation. The environment descriptions together with the assumptions correspond to a test model of the program. In this paper, we apply this modeling principle to a well known case study, the steam boiler problem which has been presented in the past. The aim of this work is to illustrate the process of building the test model and to assess the difficulty of such a process in a realistic case study. The steam boiler case study is a quite suitable problem to use, in point of both problem size and complexity, for our purposes. Taking advantage of the new features recently added in LUTESS , we show a way of defining a test model so that the testing is efficient.