Most connectivity pilots ask whether a dataset can move. This one asks a blunter question first: how much of the eFTI dataset already exists in documents the industry produces anyway, and how much has to be typed again?
The pilot takes one loading unit — a container, say — whose journey is covered by separate contracts or documents for each leg, and tries to reconstruct a single door-to-door consignment movement from them. Road first and last mile come from UN/CEFACT eCMR data; the rail leg from CIM, SMGS or CIM-SMGS; the maritime leg from a UN/CEFACT electronic bill of lading, plus FIATA forwarder bills and SMDG shipment instructions where available. Identifiers link the legs; the linked result pre-fills the eFTI dataset; a UIL or QR reference is generated; and a competent authority retrieves it through a test or simulated access chain.
The output is as much a measurement as a demonstration. The pilot records a coverage score — the share of eFTI data elements populated automatically from source data — alongside a gap and inconsistency report listing what was missing, what conflicted, and how much manual completion was needed. Its own framing is that this gives the project a measurable data-quality pilot rather than only a connectivity pilot. That distinction matters for anyone planning implementation: knowing that the pipes work says nothing about whether operators can fill them without duplicating effort.

