A Socio-Technical Resilience Framework for Platform-Based Last-Mile Supply Chains: The Interaction Between AI Technology Adoption and Human Adaptation

The last-mile supply chain concentrates multiple risks—urban congestion, demand volatility, labor instability, platform competition, and environmental constraints—making it a critical determinant of overall supply chain resilience. While AI-driven systems such as demand forecasting, routing, and dispatch optimization enhance operational stability and agility, human capabilities remain essential for exception handling, contextual judgment, customer interaction, and collaborative problem-solving. Grounded in Social-Technological Systems (STS) theory, this study proposes a conceptual framework that views last-mile resilience as a joint outcome of AI-driven and human-driven resilience. The framework highlights how platform based gig work characteristics— algorithm–rider relationship quality (A-LMX), peer network exchange (N-TMX)—moderate the complementarity between technological and human capabilities. By integrating STS principles into the last-mile context, this study contributes to supply chain resilience theory and offers managerial implications for designing, governing, and improving platform-based last-mile operations.