Software safety cases — rigorous structured arguments that some given safety-critical software system is safe enough to operate — are widely used. And there’s a lot of research, including much here at York, on how we might use them better. But there’s very little public knowledge about how software safety cases are used in practice — what actually happens on the ground when engineers, analysts and regulators think about them, talk about them, develop them, and review them . We have lots of anecdotal and informal experience, but almost no rigorously-gathered information. This means that some of our method development may be misguided and thus unhelpful or even harmful [2, 3, 4, 5].
In this project, you will study working safety-critical software developers, along with software safety engineers, analysts and regulators, in order to understand exactly what they do in practice when they learn about, develop, and review safety cases. You’ll use surveys, interviews, and (ideally) ethnographic or contextual-inquiry observation in real work environments (e.g. as done in ) in order to build a compelling process model  of how software safety-case work is done, what problems are encountered in doing that, and how practitioners deal with those problems.
(Topics you might consider include, but are not limited to — What misconceptions do practitioners have about proponent and teacher intent? What misconceptions do proponents and teachers have about practical realities? How do people learn to develop or understand and develop safety cases? Where are there disagreements (between or within groups mentioned above) as to what's good practice and why?)
Following your work, researchers (including many at York) will be able to do work on software safety cases that has better practical impact.
Social science research skills will be very valuable for this project, as will real-world industrial experience.