Novel Developments in Formal Methods for Human Factors Engineering

Abstract

Formal methods are robust tools and techniques for modeling, specifying, and mathematically proving properties about (verifying) systems. They are particularly good at both finding unexpected problems that arise from complex system interactions and proving that specific types of problems will never manifest. Formal methods have predominantly been used in the analysis and design of computer hardware and software systems. However, a growing research area within the human factors engineering community has been examining how formal methods can be used to prove whether problems exist in systems that rely on human-automation and human-human interaction for their safe operation. This symposium contains four papers by researchers who have been pushing the boundaries of where and how formal methods can be used in human factors engineering.

Task-based Automated Test Case Generation for Human-machine Interaction

“…Model checking is widely used in computer software and hardware engineering. They have also received attention from the human-machine interaction community that seeks to apply them in the engineering of reliable human-machine systems (Bolton, 2017a;Bolton et al, 2013). In particular, task behavior models can be included in larger formal verification analyses of complex systems (Bolton et al, 2013) and formal-methodsbased automated test case generation.…”

Editorial Special Issue on Computational Human Performance Modeling

“…As such, they have primarily been used in the design and analysis of computer hardware and software systems. Furthermore, a growing body of research has been investigating how they can be applied to human performance modeling to engineer safe and effective human-machine systems [20]- [22]. This special issue contains three articles, [items 5)-7) in the Appendix], that advance the use of formal methods with human performance modeling.…”

A computationally efficient formal method for discovering simultaneous masking in medical alarms

“…Even with this limitation, model checking has demonstrated its utility for a variety of applications, especially for computer hardware and software [64]. Researchers have used model checking to successfully find and correct human factors issues in automated systems [6,10,20,51,63] and medical systems [3, 4, 7-9, 54, 60]. However, outside of our previous efforts on alarm masking modeling and detection [11,[34][35][36], no work has used model checking to find safety problems associated with human sensation and perception.…”