Study unlocks AI potential in powertrain design

The project is part of the Smart Shipping Acceleration Fund (SSAF) Competition, funded by the UK Department for Transport and delivered by Innovate UK. SSAF is part of the Department’s UK Shipping Office for Reducing Emissions (UK SHORE) programme, a £206M (US$278M) initiative focused on developing the technology necessary to decarbonise the UK domestic maritime sector.

Longitude was the lead partner in the project consortium comprising the University of Birmingham, DNV, Steamology and Chartwell Marine, with support from North Star Shipping.

Longitude head of design engineering Richard Featherstone said, “This collaborative AI study marks a strategic leap in ship powertrain design. It shows that AI tools can assess multiple objective functions far quicker than a human with conventional design tools.

“However, training the AI requires a large amount of input from naval architects and equipment vendors. With a fully trained tool, removing repetitive iterative calculations will allow naval architects time for greater creativity.”

The study’s starting point was the acknowledgement that the plethora of alternative fuel choices present risks and uncertainty to shipowners, while contemporary ship design methodologies struggle to assess the multiple objective functions that an owner may wish to understand. Additionally, digital twins of inservice vessels are gathering increasing amounts of data which contemporary design methodologies struggle to interrogate.

The project therefore took the operational data from one of the North Star service operation vessels (SOVs) and used AI to assess a set of different powertrain options against three objective functions: capex, opex and emissions.

Longitude assisted the University of Birmingham in training the AI tool, applying its knowledge in ship design and alternative fuels to set parameters such as weight, space, regulatory constraints.

The outputs of the AI tool were used to develop preliminary designs of four SOV variants. The first had methanol fuel storage and a methanol to hydrogen reformer, a hydrogen fuel cell and batteries. The second had liquid hydrogen storage, hydrogen fuel cell and batteries; the third diesel and hydrogen fuel storage, with diesel internal combustion engines and hydrogen turbine plus batteries; and the fourth methanol fuel storage with methanol internal combustion engines and batteries.

The project showed that AI has both clear limitations and multiple benefits. To be truly successful, the AI tool needs to become an inhouse piece of software that naval architects can directly use themselves. Plus, training the AI tool requires a lot of input from naval architects and equipment vendors.

“With these things in place, AI can provide shipowners with cost projections from which they can base their investment decisions and better convey their intentions to the naval architect. This capability is something we are now offering to our clients to offer them greater confidence in their investment decisions,” Mr Featherstone concluded.