Various researches in artificial intelligence are conducted to understand the transfer of expertise problem. Today we perceive two major independent research activities: the acquisition of knowledge which aims to define methods inspired specially from software engineering and cognitive psychology to better understand the transfer of expertise, and the automatic learning proposing the implementation of inductive, deductive, abductive techniques or by analogy to equip the system of learning abilities. The development of a knowledge-based support system “ACASYA” for the analysis of the safety guided transport systems insisted us to use jointly and complementary both approaches. The purpose of this tool is to first, to evaluate the completeness and consistency of accidents scenarios and secondly, to contribute to the generation of new scenarios that could help experts to conclude on the safe character of a new system. “ACASYA” consists of three learning modules: CLASCA, EVALSCA and GENESCA dedicated respectively to the classification, evaluation and generation accident scenarios.