AI-Powered Simulation for Massive-Scale Transportation & Logistics Networks

Speaker:
Steve Hardy, PhD, Deloitte

The Department of Defense and its commercial partners are responsible for planning for potential events that stretch years or decades into the future. But no data to analyze exists on things that haven’t happened yet. In this session, Deloitte will discuss how they use Agent-Based Simulation, Machine Learning, and Distributed Cloud Computing to test how massive-scale, complex systems perform in novel scenarios.

Detailed Abstract: While the U.S. has been focused on counter-terrorism and counter-insurgency for the past 20 years, near-peer competitors have been focused on countering U.S. power. Consequently, these competitors now demonstrate the means and intentions to offset U.S. power, including threats to the global supply chains that are the lifeblood of the Joint Force. Studying complex systems that contribute to or impact the Department of Defense’s mission is essential, but these systems cannot be fully understood by analyzing their individual components. If a complex system cannot be studied analytically, an alternative option is to explore the system through modeling and simulation. For government and commercial partners, Deloitte uses FutureScape as a modeling and simulation platform to create digital replicas of complex, real-world systems, informed by tuned Machine Learning models, and test ways to improve them. The platform incorporates Machine Learning methods to replicate complex behaviors made by individual agents during simulations. In this presentation, we will focus on how to use FutureScape to build a controlled environment around use cases relevant to the Department of Defense and its commercial partners to answer what-if questions in an expeditious and cost-efficient manner. Two recent examples of our work with the Department of Defense and other government partners include:

• Creating a digital testbed of the bulk fuels supply chain network that enables the Department of Defense to quantify the impact of cyber and physical disruptions on fuel supply and subsequently, readiness and mobilization times.
• Modeling infrastructure disruptions with a 1,100km2 digital twin of a major metropolitan area, incorporating high fidelity AI-driven models of the electrical grid, telecom, rail, and vehicular traffic, so that a US government agency can estimate disruptions and develop mitigation strategies.

Recognizing there are significant investments being made by the Department of Defense and its commercial partners on modeling and simulation solutions, our intent is to highlight an established approach to generate data and insights about hypothetical scenarios. This will ultimately help the Department of Defense evaluate the impact of potential disruptions on Operational Plans (OPLANs) and Concept of Operations (CONOPs) and test the effectiveness of interventions and investments.