Analyzing Human Learning and Decision-Making in Heavy-Duty Fleet Management to Improve Inspection Practices

Inconsistencies in inspection outcomes, whether by human inspectors or AI systems, pose significant safety risks for heavy-duty vehicles (HDVs). Human inspectors face challenges due to complex vehicle states, while AI systems struggle with sparse historical data. These limitations contribute to critical safety issues. Brake failures cause 29% of HDV crashes, and 15.6% of crashed HDVs in 2021 had brake inspection violations. Inspectors, using tacit knowledge, determine where to focus, when to stop collecting information, and how to interpret it to make decisions. To study this process, authors surveyed 20 participants over 5 inspection rounds, evaluating 8 vehicles per round with brake condition ground truths, and applied a statistical framework to analyze participants’ feature selection and learning patterns. Results showed participants prioritized brake pad thickness (77%) and vehicle age (72%), aligning with expert-identified factors. These findings highlight the value of integrating human heuristics into AI systems for more reliable HDV assessments.