A Proposal on Discovering Causal Structures in Technical Systems by Means of Interventions

Causal Discovery has become an area of high interest for researchers. It has lead to great advances in medicine, in the social sciences and in genetics. But uptil now, it is hardly used to identify causal relations in technical systems. This paper presents the basic building blocks for in-depth research. This paper reviews established causal discovery methods and causal models. In contrast to existing surveys of this domain, we focus on the causal discovery methods using interventions. Based thereon, we propose the idea of a promising interventional discovery approach for technical systems. It takes advantage of not only direct, but also indirect causal relationships, which might improve the learning process of causal structures.