Optimus Technology Inc. on Thursday previewed a digital twin of the U.S. freight network built to model how disruptions and structural changes ripple across corridors, facilities and commodities, according to FreightWaves. The Freight Intelligence Graph, under development at the Austin, Texas-based company, is aimed at strategy and risk teams first, not the dispatchers who buy most freight software.
What the Freight Intelligence Graph models
The prototype models roughly 350,000 U.S. highway-network nodes and nearly 1 million directed road segments, FreightWaves reported. It runs on top of Optimus's proprietary freight data foundation, which maps more than 450,000 geocoded shipper and receiver roles across nearly 400,000 facility locations. The graph covers more than 500,000 distinct directional city-to-city corridor combinations.
| Metric | Value |
|---|---|
| Highway-network nodes modeled | ~350,000 |
| Directed road segments | ~1 million |
| Geocoded shipper/receiver roles | >450,000 |
| Facility locations mapped | ~400,000 |
| Directional city-to-city corridor combinations | >500,000 |
"Most market intelligence explains what has already happened," said Ed Stockman, founder and CEO of Optimus, according to FreightWaves. "We are building a model of the physical economy to address a more consequential question: What happens next, and what happens after that? A change in one market can alter capacity, economics and commercial activity several corridors away. Understanding those second- and higher-order effects is essential to planning for the future."
Hyper Predictors estimate the freight that never shows up
Underneath the graph sit what Optimus calls Hyper Predictors — specialized machine-learning models that combine shipment history, economic activity, geography, commodities, seasonality, weather and network behavior, according to FreightWaves. Their job: estimate the freight that never appears in observed data.
Toby Pasquale, head of engineering at Optimus and a former 13-year Amazon veteran in network routing, optimization and predictive transportation systems, said in the preview, per FreightWaves: "Observed freight data will always leave parts of the network unseen. Hyper Predictors close those gaps by combining specialized models, each focused on a different part of the system. Together, they let us infer likely freight flows and identify where demand, loads and capacity pressure may emerge before those patterns become obvious in historical reporting."
Hurricane scenario shows ripple effects
The company's worked example is a hurricane hitting Houston, FreightWaves reported. The storm interrupts local freight, then the effects travel: rerouted shipments, repositioned capacity, shifting fuel and route economics, and pressure in markets nowhere near the coast. Analysts inside the prototype can define changes in commodity demand, diesel prices and route conditions; the system compares a modeled baseline against the scenario to show where flow pressure builds.
What the twin will not do
Optimus draws its own boundary, according to FreightWaves. The prototype is a planning and simulation system, not a live fleet map or an ETA product. It does not present a modeled route as an executed shipment, a weather warning as a confirmed road closure, or modeled pressure as actual capacity. Four principles govern the output: verified transactions stay distinct from modeled estimates; scenarios are presented as plausible, with no single outcome called inevitable; every output retains its source and limitations; and defensible abstention. FreightWaves noted that most freight forecasting tools are not built to say they do not know.
The speculative humanoid scenario
The most speculative scenario in the preview is the one Optimus cannot ground in any observed data: humanoid robots capable of building more humanoid robots and other goods, per FreightWaves. If that capability scaled, production capacity would expand faster and sit closer to end markets. Freight would shift from long-haul finished goods toward raw materials, components and localized assembly, and facilities, inventories and transportation networks would reorganize around a different production model. No such capability exists at that scale, and Optimus is not forecasting one.
Procurement and supply chain planning implications
For chief supply chain officers and procurement directors, the stated use cases — disruption planning, network strategy, infrastructure siting and market exposure — are the decisions that shape supplier concentration, corridor risk and inventory positioning, according to FreightWaves. Access is limited for now to early design partners, and the use cases sit with strategy and risk teams, upstream of the dispatchers who buy most freight software.
The Houston hurricane example illustrates how a single event could pressure markets far from the coast. The system's design lets analysts define changes in commodity demand, diesel prices and route conditions and compare a modeled baseline against the scenario to show where flow pressure builds — a capability directly relevant to planning buffer capacity and alternate suppliers in specific corridors.
The preview's most forward-looking scenario — humanoids building humanoids — is explicitly not a forecast, FreightWaves reported, but it shows the tool's stated ambition: estimating freight that has never shown up in observed data and exploring structural shifts before they appear in historical reporting. As Stockman put it, the goal is to answer "What happens next, and what happens after that?"
Sources:
