Rotterdam turned port paperwork into one neutral data utility for the whole port

A port is a chain of independent companies — shipping lines, agents, forwarders, terminals, haulers, customs, inspectors — each holding a piece of the same shipment's data. In Rotterdam they each re-keyed it into someone else's system, and two attempts to fix this failed: the INTIS project of the 1980s died for lack of users, and the 1990s PCR and PCR-RIL projects never gathered port-community support. The port authority could build software; it could not make rivals trust a system that belonged to one of them.

The third attempt inverted the governance before writing more code. Port Infolink B.V. was set up in 2002 as a separate entity whose supervisory board seated the Port of Rotterdam Authority, Dutch Customs, Deltalinqs (the port industry association) and the companies themselves — the paper notes this representation is what let the system take a neutral position among commercial rivals. It started with import flows that Dutch Customs already wanted to automate, converted a zoo of message formats into one common XML, and ran a modular architecture so services could be added one at a time.

Adoption followed trust. By 2005 Port Infolink handled about a million transactions a month; by the end of its life it offered 24 services to some 4,500 users. In 2009 it merged with Amsterdam's PortNET into Portbase, giving the whole country one port community system — 25 services and about 1,300 clients at launch, owned 75% by Rotterdam's and 25% by Amsterdam's port authority, run not-for-profit and breaking even on transaction fees.

The compounding effect is the point: by the end of 2016 Portbase had 3,900 member companies and 14,000 users generating 82 million transactions across dangerous goods, customs, logistics and navigation services — and it had spread beyond Rotterdam to every Dutch seaport from Delfzijl to Zeeland.

Make the data utility neutral, then everyone joins

Earlier projects failed on governance, not technology: no company would feed a system its competitors controlled, so the port built a company none of them controlled alone.

Seating Customs and the industry association on the supervisory board let one system carry regulated declarations as well as commercial messages, which multiplied the reasons to connect.

A modular service architecture let the community start with one painful flow (import/customs) and expand, instead of betting everything on a big-bang launch.

Data governance was engineered early — one standard format and need-to-know visibility — so rivals could share a pipe without sharing secrets.

Not-for-profit, break-even pricing removed the fear that the utility would tax the port's own margins once everyone depended on it.

Shared infrastructure succeeds when its operator is nobody's competitor and nobody's boss. Two decades of top-down port IT failed; adoption followed neutral governance.

Portbase became permanent infrastructure: the case study records it running in every Dutch seaport, and member companies even build their own services on its platform. When Indonesia built its national port system Inaportnet, a 2024 ITS study benchmarked it against Portbase, describing the Dutch system as operationally proven in Dutch ports and noting governance practices — like two-step verification against data leaks — that Inaportnet had yet to match. What began as Rotterdam's fix for duplicate paperwork is now the reference model other countries copy.