Developed in cooperation with MOL Information Systems, the
plan is based on “mathematical optimization,” one of the fundamental
technologies of AI, the company said.
MOL operates about 100 car carriers, and in general, cargo
capacity per ship is around 5,000 standard passenger cars. In recent years,
transport and logistics patterns of automakers and other shippers have been
diversifying, and efficient vessel allocation and cargo loading are essential
to ensure the fleet to operate at peak effectiveness to meet customer needs.
As the company explained, when the vessel is calling several
loading and unloading ports, the deck and hold in which cargo is loaded can
significantly affect safety of cargo operation and its efficiency. In addition,
since the order of cargo loading/unloading and hull balance during the voyage
must be taken into consideration, it can take longer to develop a loading plan,
depending on the plan’s level of difficulty and the skills of the planner.
In this study, in cooperation with Associate Professor
Umetani, of Graduate School of Information Science and Technology, Osaka
University, two teams developed an algorithm that efficiently generates a proposed
plan from an enormous number of combinations by using mathematical
optimization. Both teams will assess the potential for practical use of the
technology, aimed at improving services through digitalization, and shortening
the required time to respond to customers when transport volume or the order of
port calls changes suddenly.
This is not the first time the company is implementing AI in
its operations. In 2016, MOL carried out a study aimed at developing the
capability to analyze data and forecast the ocean shipping market and bunker
prices with greater accuracy by using artificial intelligence (AI). More
recently, the company installed augmented reality navigation systems on 21
MOL-operated very large crude oil carriers (VLCCs).