For ShipLogic, AIS is an important source of data. So when an AIS Summit is held, we naturally want to be there

Last week, we attended the AIS Summit in Hamburg. There were some icons of the industry, such as Peter Stoyanov from AISHub and VesselFinder. But also representatives from Lloyd’s List Intelligence, Orbcomm and Luxspace, for example. In other words, people who truly have something to say.

Shadow Fleet and Data sharing

A wide range of topics were discussed — including some fascinating ideas about new business models for collecting AIS data, where contributors are rewarded for their input. It’s an interesting development that could change the way data is collected and shared, while also increasing support for open data collaboration.

The presentations on spoofing and jamming — techniques often used by actors trying to evade international sanctions (the so-called shadow fleet) — were equally enlightening. The methods for detecting such behaviour by matching radar patterns with AIS signals looked very promising. It’s a familiar and relevant issue for ShipLogic as well.

AIS and AI

Of course, one topic couldn’t be left out: artificial intelligence. AIS is, by nature, a field that lends itself well to AI applications. The solutions that were shown were largely based on LLM's. But the big question is: are today’s Large Language Models (LLMs) really the right solution for this? The examples shown, at least to me, were not entirely convincing. One of the elephants in the room remains hallucination — how can we ensure that the conclusions an LLM draws from vast amounts of data are actually correct? Or is it, for now, still mostly a clever talk box with questionable answers?

No-Nonsense

What ShipLogic needs are discrete, consistent and, above all, reliable insights from AIS data — insights that can be applied directly within our workflow. LLMs are first and foremost accessible, but not yet where they need to be.

That’s how we operate as well: if something doesn’t clearly add value, we won’t build it in — let alone promote it. But that doesn’t mean we’re ignoring AI; quite the opposite. Our technical heart beats for innovation. We continue to experiment, explore and innovate, but we’ll only introduce new technology when it truly benefits our users.

Manfred Suttorp