4.7 Article

A novel probabilistic approach to counterfactual reasoning in system safety

Journal

RELIABILITY ENGINEERING & SYSTEM SAFETY
Volume 228, Issue -, Pages -

Publisher

ELSEVIER SCI LTD
DOI: 10.1016/j.ress.2022.108785

Keywords

Counterfactuals; Causality; Bayesian network; System safety; Accident analysis

Funding

  1. National Science Foundation, United States [2045519]
  2. A. James Clark School of Engineering at the University of Maryland
  3. Marvin Roush Fellowship
  4. Ernest Lever at GTI Energy
  5. Div Of Civil, Mechanical, & Manufact Inn
  6. Directorate For Engineering [2045519] Funding Source: National Science Foundation

Ask authors/readers for more resources

This paper presents a novel probabilistic approach to counterfactual reasoning, allowing rigorous evaluation of counterfactual hypotheses in system safety through causally-sound probabilistic methods. The capabilities of this approach are demonstrated through a real-world case study on the 2018 Sun Prairie gas explosion.
Safety-critical systems cannot afford to wait for data from multiple high-consequence events to become available in order to inform safety recommendations. Counterfactual reasoning has been widely used in system safety to address this issue, enabling the incorporation of evidence from single events with an analyst's current knowledge of a system to learn from past events. However, current counterfactual methods have been criticized for making analysts prone to linearizing and oversimplifying complex events. In order to overcome these limitations, this work establishes a novel probabilistic approach to counterfactual reasoning called possible worldscounterfactuals. This methodology enables the integration of an analyst's causal knowledge about a system (in the form of a Bayesian network-based risk assessment model) with the best available evidence about an event of interest (e.g., an accident). As a result, counterfactual hypotheses, commonly used in the practice of system safety, can now be rigorously assessed through causally-sound probabilistic methods. We demonstrate the capabilities of possible worldscounterfactuals with a real-world case study on the 2018 Sun Prairie gas explosion and show how this approach can provide additional lessons and insights beyond those provided by authorities at the time of the event.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available