參考答案
I begin by ingesting real-time data from ERP, WMS, TMS, and IoT sensors into a cloud-based graph database that mirrors facilities, SKUs, lanes, and capacities. Using discrete-event simulation in AnyLogic, I then inject stochastic disruptions—such as port closures, demand spikes, and supplier defects—and measure their impacts on service, cost, and emissions. The digital twin supports policy experiments, including dual sourcing, safety stock buffers, and modal shifts. In a recent electronics case, we found that adding a strategic buffer in a nearshore DC improved service resilience by 9 percentage points during a simulated two-week ocean freight delay, at a marginal increase of 1.5% in inventory costs. The twin became a living “control tower” tool, empowering leadership to move from reactive firefighting to data-driven contingency planning.