NYK, MTI, and GRID to сollaborate on AI optimization of ship allocation plans
Nippon Yusen Kaisha (“NYK”), NYK Group company MTI Co. Ltd. (“MTI”), and GRID Inc. (“GRID”) have started developing a model using AI to optimize the efficiency of ship allocation plans of pure car and truck carriers (PCTCs), according to NYK's release.
NYK operates about 120 PCTCs, the largest fleet among shipping companies worldwide. The ship allocation plan, which determines which port a vessel will sail to start the next voyage once the current voyage ends, is generally formulated by a skilled person after consideration of various conditions such as cargo demands, ship schedule, vessel type, and ship loading capacity. NYK currently utilizes its own in-house ship allocation planning system but has faced difficulties dealing with various decision-making factors and situations that can change from moment to moment. In addition, as efforts for decarbonization accelerate within the shipping industry, operating these 120 PCTCs as efficiently as possible has become an imperative issue to reduce greenhouse gases (GHGs) emitted from ships.
To address these topics, NYK, MTI, and GRID are now collaborating to optimize the ship allocation plans for NYK’s PCTC fleet. These three parties aim to build an optimization model for ship allocation plans by combining NYK’s know-how in formulating ship allocation plans, MTI’s simulation technology in ship operations, and the AI technology of GRID, a technology venture specializing in social infrastructure.
In this collaboration, digital twins and the latest machine learning technologies will be introduced, in addition to the mathematical optimization technology that has been used in NYK’s own in-house system. Not only building of the optimization model but also application development will be included in the scope of this collaboration.
NYK, MTI, and GRID are aiming to reduce GHGs from ships, as well as improve the efficiency of the planning process, and the three companies will continue to optimize future models and systems. Trials of this system are planned for June 2022, with full operation targeted for 2024.
