Journal of System Safety
Not a member yet
238 research outputs found
Sort by
From the Editor's Desk: AI Tools for System Safety Analysis?
No full textIt was not long ago that there were reports of students using AI applications to write assignments for school. Now it is common for engineers to use AI applications to help with engineering analyses and even to generate software. Some people are concerned as there are cases where AI makes up answers to questions (referred to as hallucinations). There are also situations when it is not clear how the model arrived at a particular answer. To be fair, I seem to remember some analyses generated by full blooded human beings that seemed to have made up answers and difficult to understand conclusions
In Memoriam: Dr. Loan (Joan) Pham
Get PDFOur distinguished colleague, Dr. Loan (Joan) Pham, passed away suddenly on August 16, 2023. This is a tragic and overwhelming loss to all who knew Joan, but she and her work will be remembered throughout the fields of system safety and aviation safety
Reduction of Normalization of Deviation (NoD) Using a Socio-Technical Systems Approach
Get PDFNormalization of deviation (NoD), also known as normalization of deviance, is the process in which deviations from correct or proper decisions, behaviors, or conditions important for safety insidiously become the accepted norm over time. NoD is a common, risky, yet elusive issue causing or contributing to numerous accidents in multiple industries. Effective reduction of NoD is therefore a major opportunity. Approximately 10 years ago, Boeing developed a general systemic model of NoD based on a socio-technical systems approach. It is a representation of how multiple internal and external factors inherent to socio-technical systems interact in a dynamic fashion leading to NoD. It holistically captures the essence and complexity of the problem. The model has been shared across Boeing and with three customer airlines of Boeing. Specific systemic models of NoD associated with specific problems were developed based on the general systemic model. Subsequently, NoD awareness training, methods, tools, processes, and solutions based on those models have been developed. They were provided and/or used to improve workplace safety at Boeing and aviation safety at one of the three airlines. All the efforts have resulted in unprecedented insights, and some have seen significant reduction of NoD, NoD-related incidents, and NoD-related safety risks
Review of the Latest Developments in Automotive Safety Standardization for Driving Automation Systems
Get PDFThe ISO 26262: Functional Safety – Road Vehicles Standard has been the de-facto automotive functional safety standard since it was first released in 2011. With the introduction of complex driving automation systems, new standardization efforts to deal with safety of these systems have been initiated to address emerging gaps such as the human/automation roles and responsibilities in the presence/absence of the driver/user, the impact of the technological limitations and the verification and validation needs of automation systems to name a few. This paper highlights some of these gaps and introduces some of the latest developments in automotive safety standardization for driving automation systems
From the Editor's Desk: Cognition
No full textWhen accidents occur the question “What were they thinking?” is often asked by those of us who are investigating the situation. When we do system design, especially the system safety aspects, we often consider the cognitive process (and errors associated with this process) of the operator. We often also wonder about the thought processes of our colleagues, our management and probably should also wonder about ourselves. The work of Nobel Prize winner Daniel Kahneman [ Ref 1] has stimulated some interesting discussion in the last several years. The main finding of Kahneman is that our minds are susceptible to systematic errors. I will enumerate a few examples that I find particularly interesting
TBD
Get PDFI want to tell you a story about an encounter I had at a hotel bar in Lancaster California. I appreciate that at first it doesn’t appear to have anything to do with System Safety. Trust me, I think you will agree that perhaps there is an important lesson for us and the Society
From Our Readers
Get PDFLetter to the Editor:
In the Charles Hoes article “TBD regarding Risk Assessment” the sample risk assessment matrix and his explanation of how the chart can be used to assign “risk levels” is on par with the basics of the risk management process generally used on programs
Proposing the Use of Hazard Analysis for Machine Learning Data Sets
Get PDFThere is no debating the importance of data for artificial intelligence. The behavior of data-driven machine learning models is determined by the data set, or as the old adage states: “garbage in, garbage out (GIGO).” While the machine learning community is still debating which techniques are necessary and sufficient to assess the adequacy of data sets, they agree some techniques are necessary. In general, most of the techniques being considered focus on evaluating the volumes of attributes. Those attributes are evaluated with respect to anticipated counts of attributes without considering the safety concerns associated with those attributes. This paper explores those techniques to identify instances of too little data and incorrect attributes. Those techniques are important; however, for safety critical applications, the assurance analyst also needs to understand the safety impact of not having specific attributes present in the machine learning data sets. To provide that information, this paper proposes a new technique the authors call data hazard analysis. The data hazard analysis provides an approach to qualitatively analyze the training data set to reduce the risk associated with the GIGO
