Safer, more efficient, and more environmentally friendly. With the help of AI and machine learning, cargo handling on oil and chemical tankers can be optimized. That is shown by the results from the ACHT 2.0 project, funded by the Swedish Transport Administration and led by RISE.

The complexity of today’s cargo operations on tankers is high. These operations are both complex and risky: hazardous cargo must be handled in parallel with ballast and inert gas systems, while the vessel’s stability must never be compromised. One challenge is that these separate systems do not communicate with each other. This means that the bridge officers become manual intermediaries—a bottleneck that creates inefficiency while also increasing the risk of mistakes.

“A lot happens at the same time during loading. The ballast must be treated and pumped out while cargo is taken in. At the same time, the tanks must be kept free from explosive environments, which requires inert gas that is then ventilated out. And in the middle of it all, the chief officer becomes an information hub—the phone rings from shore with questions about schedules and capacity. In short: it’s a situation that demands concentration on many fronts simultaneously,” says Fredrik von Elern at RISE, who has led the ACHT 2.0 research project.

The project builds on the pre-study ACHT (Automated Cargo Handling on Tankers). The pre-study not only mapped the systems used for cargo handling and how they are controlled, but also showed that it is entirely possible to develop an integrated cargo handling system with today’s technology.

In ACHT 2.0, the next step was taken: a prototype was developed where AI and machine learning are used for cargo planning.

“We collected data from a large number of voyages carried out by Terntank’s vessels. After that, Kockumation and Combine developed a planning tool. It can generate proposals for cargo planning even for voyages with multiple stops.”

The biggest challenge, says Fredrik von Elern, was the next step—developing an autopilot system that can independently manage loading and unloading.

“The problem was that we lacked access to the real-time data required. However, we have now been promised support from a system supplier who can provide the data. This means we want to take the project further into ACHT 3.0, provided that we secure funding.”

If that happens, an autopilot system will be created—something Fredrik von Elern is convinced of.

“Bureau Veritas has been an important partner in the project, contributing invaluable knowledge about risk management, safety, and regulations. The report clearly shows the requirements that must be met for an autopilot system to be implemented. It is an extensive task that requires close collaboration with Bureau Veritas and other authorities. By integrating their expertise into the development, there is an opportunity not only to meet, but actually to exceed today’s safety and regulatory requirements.”

The research project ACHT 2.0: Proof of Concept of Automated Cargo Handling on Oil and Chemical Tankers has been carried out by Diewertje Dekker, Sofia Hörberg, Pablo Correa Gomez & Sofya Krainova-Parfenovich (Combine), Niklas Dahlberg (Metstech), Fredrik Ekelöf (Kockumation), Nicole Costa & Fredrik von Elern (RISE).