Smart Port starts with the processes

At the Port of Tallinn, cars and lorries can pass through the automated check-in lanes in around two seconds, according to Kadri Haufe, Managing Partner of the maritime consultancy Hafring. Before the changeover, the average waiting time was around two minutes. There is more to such progress than just new technology: “A smart port is a process, not a product,” explained Haufe at the “Port in Transition” expert forum in Lübeck. It is crucial that data, once recorded, is utilised for various stages of the process from the port gate right through to loading.
The Lübeck Port Authority (LHG) facilitated the exchange of such practical experiences in the run-up to the German-Finnish Port Day: “More than 15 ports are represented here today,” said LHG Managing Director Prof. Sebastian Jürgens at the opening. “This is valuable for all of us because it allows us to share experiences from the various ports and thereby reach a particularly wide audience.” Robert Kiehn, Head of the IT Department for Business Apps, HR & Finance at LHG, moderated the forum, which was attended by over 80 participants.

Integrate first, then invest

Haufe recommended that port operators follow a set sequence for process optimisation: utilise existing technology, integrate existing systems, and only then invest in missing functions. If, for example, registration numbers, vehicle dimensions and booking data were recorded during vehicle processing before the port gates, they could also be used during check-in, traffic management and loading. “The prerequisite is that port authorities, terminal operators, shipping companies and other stakeholders share their data and coordinate their processes,” said Haufe.

First, it must be established which key performance indicator is to be improved – such as the time taken from the gate to the ramp. Sometimes, better signage or clearly marked lanes are sufficient. Automation and AI forecasting should only follow once the fundamentals are in place.

Haufe illustrated the potential benefits with further figures: at the Port of Tallinn, 75 per cent fewer full-time staff are required for check-in and traffic management in the staging area. In Rosslare, Ireland, 40 hours of administrative work are saved each week. Another potential benefit lies in more accurate billing: if vehicle dimensions do not match the booking, digitally recorded data could prevent revenue losses.

Applications from Lübeck
Sabine Welchert and Maximilian Mang from Lufthansa Industry Solutions presented specific applications: at LHG’s Skandinavienkai, data on incoming vehicles is analysed to identify traffic peaks and long waiting times in limited parking areas.
Another project focuses on incorrectly parked trailers. If these have to be located first, this results in additional journeys and wasted time. With the aid of cameras, identification codes and number plates are to be recorded, and the actual parking spaces displayed on a digital site plan. Welchert and Mang presented a trial concept, simulations and a draft dashboard for this purpose. Another focus was on the docking of ships: radar and laser sensors, combined with wind and current data, are intended to map the approach to the berth and thus ensure safe manoeuvring.

There is no alternative
The more port processes are digitally interlinked, the more operations depend on a functioning IT system: Andreas Reichel, Chief Technology Officer at the public IT service provider Dataport in Hamburg, therefore focused on the consequences of serious cyber-attacks. Alongside prevention, detection and response, a fourth task is crucial: “The fourth pillar is resilience – in other words, getting a paralysed system up and running again.”
In an attack scenario outlined by Reichel, attackers gain access rights, destroy backup copies and cripple end devices and networks. Communication channels could also fail as a result. A second data centre alone does not provide sufficient protection if both sites use the same vulnerable software components. What is needed are technically independent fallback systems and well-rehearsed procedures for alerting and recovery.
Such safeguards begin with day-to-day operations, as Haufe had previously emphasised: if, for example, automatic number plate recognition fails, it must still be possible to enter the premises by alternative means.

AI must prove itself in everyday use
Similar standards apply to artificial intelligence: “Specialised AI tools are particularly well suited to recurring and standardised tasks,” explained Johann Bergmann from the Fraunhofer Centre for Maritime Logistics and Services (CML). Bergmann reported on projects relating to staff scheduling, documentation and automation. An experiment involving the automatic operation of locking pins on container wagons demonstrated just how challenging practical implementation can be: sunlight and reflections initially significantly impaired image recognition.
The forum’s conclusion: it is only when digital technology is deployed within coordinated processes that it represents progress with measurable benefits – and a plan in the event of a fault.