AI-Augmented Resilience Engineering: Integrating Generative AI with Simulation and Optimization in High-Consequence Systems

Abstract:

Healthcare systems and disaster response environments operate under profound uncertainty, behavioral heterogeneity, and tightly constrained resources. Modeling these systems requires frameworks that capture both structural dynamics and human decision-making under disruption. While simulation and optimization provide formal analytical structure, they often struggle to represent adaptive behavior and fragmented institutional knowledge. Conversely, generative AI offers powerful reasoning capabilities but lacks structural grounding when deployed independently.

This talk presents a research program that embeds large language models within hybrid system dynamics, agent-based, and optimization architectures to create AI-augmented decision-support systems. Applications include bidirectionally coupled system dynamics and agent-based models for opioid use disorder policy analysis, multi-persona AI agents for disaster response simulation, and AI-assisted maintenance planning in reliability-critical infrastructure. Across domains, optimization methods support calibration, work packaging processes, and policy sensitivity analysis.

This work advances a structured framework for integrating generative AI into simulation and optimization environments, enabling interpretable, policy-relevant, and computationally grounded systems that enhance, rather than replace, human judgment in high-consequence socio-technical settings.

About the Speaker:

Gerald Jones is a Ph.D. candidate in Industrial and Systems Engineering at the University of Tennessee and a Graduate Research Assistant in the Applied Systems Lab. His research integrates artificial intelligence, simulation modeling, and optimization to improve decision-making in mission-critical systems. His work spans AI-assisted maintenance planning for high-reliability facilities, hybrid system dynamics and agent-based models for opioid use disorder policy analysis, and multi-persona AI agents for disaster response simulation. With academic foundations in electrical engineering, computer science, and systems engineering, his research focuses on building resilient, AI-augmented decision environments for complex socio-technical systems.