Cargo Security Improvement Actions in High-Risk Urban Logistics Using a Paraconsistent Expert System: A Comprehensive Decision-Making Framework

Background: Cargo theft remains a persistent and critical challenge in last-mile logistics, particularly within high-crime urban peripheries such as those in São Paulo, Brazil. These incidents compromise transportation security, disrupt supply chains, and impose considerable financial and operational burdens on logistics providers. Traditional security measures often prove inadequate in addressing the dynamic and complex nature of such criminal activities, highlighting the need for adaptive, intelligent decision-support systems capable of managing uncertainty, contradictions, and real-time risk assessments. Methods: This study synthesizes findings from three prior investigations to propose a comprehensive decision-making framework for cargo security, underpinned by a Paraconsistent Expert System (PES) based on Paraconsistent Annotated Evidential Logic Eτ. Employing a mixed-methods approach, the study integrates qualitative evaluations from logistics professionals with quantitative analyses of six targeted security interventions. A panel of nine experts-ncluding logistics managers, operational coordinators, and frontline delivery personnel-assigned degrees of certainty and uncertainty to each intervention. The PES processed these inputs to reconcile conflicting assessments and support strategic decision-making under conditions of informational ambiguity. Results: The analysis revealed that five of the six evaluated strategies-such as GPS tracking, armed escorts, optimized delivery scheduling, and the recruitment of local drivers-substantially reduced cargo theft incidents. The use of the Paraconsistent Expert System enabled more accurate risk assessments and enhanced the responsiveness of security operations. Notably, security-related costs decreased by 58% in Arujá and 75% in Cajamar, two critical logistics hubs in the São Paulo region. Conclusions: The proposed framework demonstrates the viability of a scalable, data-driven model for improving cargo security in high-risk urban logistics. By incorporating Paraconsistent Logic Eτ, the system offers a robust analytical tool for managing contradictory and uncertain data, thereby strengthening risk mitigation and operational resilience. The findings advance the literature on AI-supported decision-making in logistics and offer actionable insights for industry practitioners, policymakers, and security agencies seeking to enhance cargo protection in vulnerable urban areas.