Systems Theoretic Process Analysis of Sociotechnical Systems

Abstract

The safety and success of complex modern systems, such as hospitals, aircraft, or software, depend on their ability to integrate people and technical components. For example, doctors must be able to use their computerized surgical tools to treat their patients successfully, airplane pilots must be able to operate the required controls for takeoff and landing, and regulators must be able to interpret the data they receive to make critical decisions. However, designing systems that facilitate safe interactions between humans and technology is not a simple task. System designers must consider not only the constraints of the technical components but also human requirements throughout the entire system. However, accidents in modern systems continue to prove that more work is needed to identify and prevent unsafe interactions between humans and technology Systems Theoretic Process Analysis (STPA) is a hazard analysis methodology based on systems theory that has been used to improve system safety in various industries, including healthcare, aviation, nuclear power, and automotive design. However, if hazard analysts using STPA lack significant expertise in human factors engineering (HFE), they may be unable to thoroughly and rigorously identify critical unsafe interactions. This thesis presents a process for utilizing HFE to improve the results of STPA analyses on sociotechnical systems. In particular, the process focuses on the thorough identification of causal scenarios in sociotechnical systems by incorporating relevant human factors concepts. The process allows analysts without significant training in HFE to improve their ability to identify useful scenarios for humans in their system. The effectiveness of the improved process is demonstrated using a healthcare case study on over-the-counter clinical laboratory tests in the United States. By establishing a process for non-HFE experts to use when conducting STPA analyses, more systems can be developed that enhance human performance rather than increase conflict between humans and the engineered system.