Challenges in Future Mathematical Modelling of Hierarchical Functional Safety Control Structures within STAMP Safety Model

¹ Management & Safety Engineering Unit, Engineering Faculty, Ben-Gurion University, Israel
² Department of Mechanical Engineering, Engineering Faculty, Ben-Gurion University, Israel
³ NRC-Negev, Israel

^* Corresponding author: lbaror@yahoo.com

Abstract

In the STAMP model, based on control theory, the control relationships between various system elements enforced by the closed Control Loops (CLs) are logical and functional. A literature survey emphasized the fact that for the moment STAMP and its main tools STPA and CAST are not associated with any numerical tools. The main rationale of our work is to understand whether STAMP matches to be a quantitative model. Furthermore, in a case that we find that numerical tools can be used in STAMP, we intend to bridge the gap between the logical-functional approach in STAMP and any of the suitable quantitative approaches applied in Engineering Control Theory (ECT). As a first step, a literature comparison was performed between the basic control parameters existing explicitly at the moment in the STAMP model, and those well known in the literature of ECT. The results reveal that there are many similar terms, especially related to conceptual and general definitions. However, we have observed that there are also basic quantitative parameters from ECT which are not yet referred to in STAMP as quantitative safety evaluation parameters. Another main finding is an inherent difference in various ECT related parameters and the CLs at the various hierarchical levels. ECT was originally developed to deal with physical systems. Thus, any machine related internal control loops within the lower-physical level of a Sociotechnical System (STS) can be directly addressed with quantitative methods from ECT. However, most of the human-machine interactions in the lower levels and the human and societal controls in the higher levels are at the moment not suitable for those methods. We assume these ECT parameters may have an important role in designing and examining systems safety and hence we suggest, should be integrated into STAMP model, in purpose to be able to enhance systems safety.