How the twin works

A network graph of U.S. freight

The twin represents freight as a graph. Nodes are the places where freight starts, stops or changes hands: plants, warehouses, ports, rail ramps, distribution centers and carrier terminals. Edges are lanes: the routes that connect them by truck, rail, water and air.

Each lane carries what a planner already thinks about: what moves on it, which carriers serve it, how much capacity it has and what it costs. Shippers and carriers are tied to the facilities and lanes they use, so a change in one place can be traced to the companies it touches.

Data inputs

The model is built to combine public data about the network, such as infrastructure, trade flows, weather and fuel prices, with signals about freight demand and capacity. For design partners, it is designed to take in the lanes and facilities they run, so the twin reflects their part of the network.

Simulation

A scenario is a change to the graph: a facility closes, a lane gets faster or slower, demand moves to a new region, a cost goes up. The twin is built to apply the change where it lands, then follow it outward through the lanes and facilities connected to it, step by step over time.

Because freight is a network, the effects are rarely local. Capacity that leaves one region is missing from another, and a lane that gets cheaper pulls freight from its neighbors. The simulation is built to show those second and third effects, not only the first.

Outputs for planners

Results are described in the terms planners use: which lanes and regions would tighten or loosen, where capacity would run short, which way rates would move and how long the effect would last.

Every result is a scenario, not a forecast. The twin is a tool for comparing options before a change arrives.