Arc-level reliability integration in connected autonomous vehicle emergency routing: a multi-objective framework with critical service quality threshold analysis
Academic Article in Scopus
Critical infrastructure disruptions translate into degraded emergency medical services when routing systems are blind to arc-level reliability. This paper formulates the Reliability-Aware Connected Autonomous Vehicle Routing Problem with Fuzzy Time Windows (CA-VRPFTW-NR) as a three-objective mixed-integer program that simultaneously minimizes priority-weighted patient dissatisfaction and cumulative route unreliability and maximizes priority-weighted service coverage. A single arc reliability parameter is embedded at four mechanistic levels of the solution process: travel-time computation through a reliability-adjusted Bureau of Public Roads (BPR) function, route feasibility filtering, solution repair, and direct objective evaluation, ensuring coherent reliability-awareness throughout optimization rather than treating infrastructure degradation as a post-hoc constraint. A quantum-inspired genetic algorithm (QiGA) is proposed as the primary solver, with four benchmark metaheuristics and three multi-objective comparators evaluated under reliability conditions. Computational experiments across multiple network sizes, topology types, and failure patterns identify a critical reliability threshold below which service quality collapses nonlinearly and cannot be recovered through fleet expansion or algorithmic improvement alone. Hub-node failures produce greater service degradation than random outages at equal mean reliability, with implications for infrastructure hardening prioritization. Multi-objective analysis shows that Pareto decision flexibility contracts at the same critical threshold, making robust optimization most valuable when networks are most stressed. The framework is applied to the 2025 Iberian Peninsula blackout, where the four blackout scenario points map onto the predicted service-quality phase diagram without recalibration, indicating that the findings extend to a real infrastructure failure event. The source code, simulation metadata, and case study data are available at https://doi.org/10.5281/zenodo.21800619 .
