The Container Disappears Between Ship and Rail

A single shipping container moving from factory to warehouse may pass through 5–10 different carriers (ocean, rail, truck, barge), 3–4 customs jurisdictions, and dozens of intermediaries (freight forwarders, port authorities, customs brokers, terminal operators). Each segment uses different data formats, identifiers, and communication protocols. There is no universal interoperability standard that tracks a container continuously across all transport modes. The result is that at any given moment, ~30% of containers in transit have uncertain status, leading to terminal congestion, empty container repositioning waste, and supply chain opacity.

Global container trade moves ~800 million TEU/year (~$14 trillion in goods). Empty container repositioning alone costs the shipping industry ~$20B annually, driven largely by poor visibility into container location and availability. Port congestion — exacerbated by the inability to predict container arrivals across modes — costs the global economy an estimated $50B/year. The COVID-19 supply chain crisis demonstrated that container tracking opacity amplifies disruptions: carriers, ports, and shippers all made suboptimal decisions because they couldn't see the system state. Despite decades of digitization, the typical international shipment still generates 36 paper documents and involves 200+ data fields exchanged between parties.

Multiple standards exist but none has achieved universal adoption: EDI/EDIFACT (dominant in maritime, limited penetration in trucking), GS1 EPCIS (retail supply chain, weak in shipping), DCSA standards (carrier-led, incomplete mode coverage), UN/CEFACT reference data model (comprehensive but complex). Blockchain-based platforms (TradeLens by Maersk/IBM, Global Shipping Business Network) promised neutral interoperability but failed commercially — TradeLens shut down in 2022 — because competing carriers wouldn't share data on a competitor-owned platform. IoT container trackers (GPS, cellular) provide location data but don't solve the document/status interoperability problem. The fundamental barrier is not technical but institutional: each participant in the logistics chain has invested in proprietary systems and has weak incentives to adopt a standard that would make their services interchangeable.

A federated data architecture — where each party retains control of their data but publishes events (container loaded, departed, arrived, cleared) in a common format to a shared event bus — could provide visibility without requiring parties to share proprietary operational data. The EU's eFTI (electronic Freight Transport Information) regulation (effective 2025) mandates this approach for European transport but covers only EU-internal movements. Extending this model globally requires IMO, WCO, IATA, and UIC to align their data models — a coordination challenge that has been attempted repeatedly without success. Alternatively, large shippers (Walmart, Amazon, Maersk) could force adoption by requiring their carriers to publish events in a specified format, creating de facto standards through market power.

A team could map the data flows for a specific trade lane (e.g., Shanghai to Los Angeles via rail to Chicago) and identify exactly where data handoff failures occur and what information is lost at each transition. Alternatively, a team could prototype a lightweight event-publishing API that translates between two major freight data standards (e.g., DCSA and EDIFACT) and test it with a freight forwarder handling real shipments. Skills: logistics, data engineering, API design, systems analysis.