Fuzzifying the Operational Design Domain:A Scalable Situation Hyperspace Generation

Underground mine environments present complex conditions for autonomous systems, including narrow tunnels, variable lighting, and dynamic obstacles. In the design phase, ensuring the safety of critical drone systems under the vast and uncertain Operational Design Domain (ODD) of an underground mine is practically impossible through exhaustive testing alone. Our work proposes a fuzzy logic-based approach to focus testing by constructing a hyperspace from the ODD that prioritizes high-risk situations to test at the design phase efficiently. We demonstrate how this approach prioritizes critical test cases and minimizes redundant evaluations of low-risk situations. The resulting framework enhances safety testing in mining operations and can be extended to other safety-critical systems facing similar challenges of scale and uncertainty. The proposed fuzzy-logic based approach reclassifies many test cases that would be labeled as high risk under a disjoint method into medium-risk categories by allowing gradual transitions between risk levels, enabling more fine-grained risk assessment and more effective risk-based test prioritization while maintaining coverage of hazardous situations.