AI Case Study
Orient Overseas Container Line Limited (OOCL) looking to optimise shipping operations with deep and reinforcement learning
Orient Overseas Container Line Limited (OOCL) has announced a partnership with Microsoft Research Asia (MSRA) on a 18 month R&D project. Its scope is to apply deep learning and reinforcement learning in the company's shipping network operations in order to optimise them. According to the company's CIO, such optimisation could save the shipping giant USD10 million in costs per year.
"Shipping giant, Orient Overseas Container Line Limited (OOCL), has come together with Microsoft Research Asia (MSRA) on a digital transformation journey to identify, manage and overcome all sorts of operational uncertainties and create efficiencies.
OOCL has a proud track record of adopting new technologies with real business impact. It has fully embraced a hybrid cloud infrastructure with auto-switching and auto-scaling throughout its business and ML for several years. And, it has more than 1,000 developers in San Jose in the United States, as well as Hong Kong, Shanghai and Zhuhai in China, and Manila in the Philippines.
It regards AI as key to its future, and so it has turned to MSRA – Microsoft’s fundamental research arm in the Asia Pacific, which is leading in the development and application of AI. Recently, the two sides spent 15 weeks together optimizing OOCL’s existing shipping network operations – a task that Steve Siu, Chief Information Officer of OOCL, estimates will save the shipping company around USD10 million in costs annually.
Following that success, they have announced a new 18-month research and development partnership to apply deep learning and reinforcement learning in shipping network operations."
Research; results not yet available
"Thousands of cargo ships ply the oceans and keep world trade moving ahead every day. But poor weather, congested ports, equipment breakdowns, and mishaps of all kinds make for anything but smooth sailing. An unexpected delay for just one vessel can sometimes cascade into a logistical nightmare for an entire fleet with schedules thrown out of kilter across multiple ports and trading hubs – impacting the flow of globalized supply chains.
Adding to the physical challenges are fluctuating trade volumes among economies: A vessel that sails off in one direction with containers crammed full of goods can too often sail back unprofitably empty.
Imagine using the Internet of Things (IoT) to constantly collect data on all voyage-impacting factors, both on and off a ship. Predictive analytics through artificial intelligence (AI) would lead to operational adjustments and decisions to keep vessels moving in the most cost-effective ways.
Such insights could, for instance, help a captain save on fuel by optimally maintaining and varying speed to port. Similarly, a course or balance route time could be amended to avoid anticipated port congestion or deteriorating weather. Operational flexibility and co-ordination could be introduced across multiple vessels to smooth out the transfer of cargo across multiple points, saving both money and time. Logjams could be avoided. Under-utilized routes and schedules could be recast and made more profitable."