Disruptions in the agricultural food product (AFP) supply chain pose significant risks to economic stability and food security in the U.S. Extreme weather events, labor shortages, geopolitical conflicts, and market volatility can create cascading failures in food distribution networks, leading to increased costs, shortages, and inequitable access to essential goods. This proposal aims to develop a data-driven disruption-response framework that integrates parametric modeling, predictive analytics, and real-time monitoring to enhancethe resilience and efficiency of AFP supply chains.

This project expands the classical resilience triangle and uses the double logistic function to characterize AFP supply chain recovery dynamics across production, transportation, and consumption. A predictive modeling framework estimates supply chain service levels based on disruption characteristics, leveraging multivariate regression and machine learning to enhance forecasting accuracy. The research also explores the overlapping effects of simultaneous disruptions, employing additive and weighted interaction models to capture real-world complexities. Sensitivity analyses will further identify critical thresholds where disruptions become significant. The study develops a real-time disruption monitoring and alert system to translate research findings into actionable solutions. This system will employ natural language processing (NLP) and anomaly detection to track global and regional events affecting AFP supply chains, providing early warning alerts when disruptions exceed critical thresholds.

Aligned with the USDOT Research and Development Strategic Plan and this RFP, this research delivers data-driven insights, predictive capabilities, and practical decision-making tools to support more sustainable, adaptive, and resilient supply chains, strengthening the U.S. economy and ensuring reliable access to essential goods. The methodologies proposed are scalable to other high-velocity supply chains (e.g., pharmaceuticals,energy, and manufacturing).