S13.5 · Transportation, Logistics & Mobility

Rail Transportation

Six fixed-infrastructure Class I railroads using predictive maintenance and scheduling AI to improve operating ratio, not market structure.

S13.5

What is on this page. Market structure, and how AI is reshaping this segment. Ownership, buyer universes, transaction comparables and deal-timing analysis are maintained privately by El Dorado Capital and are not published.

US Class I railroads generate approximately $80B in aggregate revenue (Association of American Railroads data) across six carriers that together control the vast majority of US freight rail ton-miles — Union Pacific at roughly $24B, BNSF at roughly $23B, CSX at roughly $14B, Norfolk Southern at roughly $12B, plus a combined operator running a single North American network across the US, Canada and Mexico. No segment in this sector has higher barriers to entry: the infrastructure is fixed and the approvals are regulatory. That frames what AI can be here — a margin tool working a fixed asset base, not a force reshaping who competes.

Market structure

The growth story is a mix shift, not a volume story. Overall volume growth is low-single-digit, with intermodal traffic the primary growth vector against flat-to-declining coal and merchandise carload volume — a shift that favors carriers with strong intermodal terminal networks over those still weighted toward legacy bulk commodities. North America is the concentrated developed market, and a single-line network now spans the US, Canada and Mexico, giving shippers continuous cross-border service without an interchange handoff. For nearshoring-driven Mexico-US freight, that is the structural advantage pulling volume onto rail rather than truck.

For the six Class I carriers, barriers to new entry are extreme: rights-of-way are fixed and effectively non-replicable, and Surface Transportation Board oversight gates rates and service commitments. Once scale is reached, the economics are high-fixed-cost and high-margin — operating ratios in the 60s are typical — supported by long-duration shipper contracts. Upstream sit locomotive and rolling-stock manufacturers and rail-car lessors; downstream, bulk shippers directly, and intermodal and drayage trucking operations at the terminal interface.

How AI is reshaping this segment

Both of AI's primary levers in rail — predictive maintenance and network-scheduling optimization — work the same target: operating ratio. Better maintenance scheduling reduces unplanned downtime and asset-failure cost. Better network scheduling improves utilization across a fixed set of locomotives, cars and track. Every dollar either lever produces drops to margin on an unchanged revenue line, and neither touches the fixed-infrastructure moat that defines competitive position — that moat is a function of owning the rights-of-way, not of software.

Of the two strategic builds in motion, continued single-line cross-border service captures freight that would otherwise require a truck-to-rail-to-truck handoff or move by truck entirely — genuinely new volume for rail. Precision-scheduled railroading investment defends rail's cost position against trucking competition on lanes where the two modes overlap. Both are already under way, alongside predictive-maintenance deployment across the major carriers. Rail's AI adoption is mature relative to the rest of the sector, and for a structural reason: a fixed-asset base at this scale rewards marginal efficiency gains more than almost any other business model does.