Fathom World

Mapping the transformation of shipping and the oceans

Digital & ElectronicsFeatured

We can predict performance today with artificial intelligence

Artificial intelligence (AI) is on the up but there are still “significant digital leaps” ahead if maritime players want to tap the potential benefits it promises. Brian Dixon looks at what needs to be done.

There can be little doubt that the use of AI and its subset of machine learning (ML) is on the increase, with its upward adoption curve within the global maritime sector only likely to steepen over the coming years.

But according to some experts, the advances in digitalisation can already unlock opportunities with AI in the maritime industry and enable the industry to become smarter, more efficient and more sustainable.

Peter Sarlin is co-founder and CEO of Silo.AI, a three year old Helsinki-based company which calls itself the largest private laboratory promoting artificial intelligence in a number of different industries.

“Global drivers such as climate change and a continuous strive for efficiency but also advances in technology are pushing the industry towards smart ports and autonomous vessels,” he said, a reference to one of the company’s clients, Awake.AI, another Finnish company which is actively developing and delivering solutions for autonomous shipping.

Awake.AI, is currently building an open platform and ecosystem for smart ports and autonomous shipping. Noting that AI and ML will accelerate this transition in port call optimisation, Jussi Poikonen, Awake.AI’s co-founder and vice-president of AI and analytics, stated: “We believe that AI and machine learning will help in daily-task automation and operation optimisation so that people can focus on (better-informed) decision making and other developments in maritime logistics.” In this light, Poikonen envisages artificial intelligence and machine learning as a means to reducing human error and inaccuracy, its deployment in the modern maritime sector complementing rather than replacing human staff.

“The main driver really is the fact that the maritime industry is now ready for digitalisation and there is an increasing number of investments made in developing the maritime logistics’ digital infrastructure. Digitalisation and electrification are the two mega trends in the industry, meaning that new technologies such as AI will be adopted by the industry,” added Awake.AI expert, CEO and co-founder Karno Tenovuo.

However, it’s not all plain sailing for this particularly technology. “Silo.AI works with transforming many traditional industries with AI, and as with many others, the maritime industry is impacted by low digital maturity and an overall skill and competence gap in unlocking opportunities with AI technologies,” Sarlin stated.

With digitalisation arguably lagging behind other sectors, actors in maritime, Tenovuo noted, “rarely have the needed tooling and platforms to stand on, covering necessary aspects like digital applications, data standards, [application programming interfaces (APIs)], data sharing and cloud services. In building intelligent solutions for vessels, ports and logistics alike, they all oftentimes still need significant digital leaps when moving from pen and paper, analogue devices and manual labour to AI-powered solutions.”

And when it comes to building intelligent solutions, Awake.AI has spent the last year collaborating with Silo.AI on the development of two machine learning systems designed to improve situational awareness at ports, with the first employing an ML model to improve precision when different stages of the port call process occur. “The developed solution,” Tenovuo said, “reduces significantly the error margins in the port call predictions, allowing port operators to do better resource planning and improve operational efficiency.”

Meanwhile, the second project saw the two companies developing an ML algorithm that, coupled with computer vision, analyses and monitors cargo and vehicles within a port in real time. This system, Awake.AI’s Poikonen said, not only automates challenging manual analysis, but also speeds up cargo logistics planning and improves the detection of potential exceptional situations. Both solutions, he also said, have been developed together with end users and Silo.AI and have been piloted in real environments.

Looking to the future, Awake.AI and Silo.AI intend to continue their collaborative efforts to develop predictive analytics and other ML applications for ports and their logistics chains. In addition to this, Silo.AI has also worked on various other logistics projects, including parcel delay prediction for the Posti Group (the Finnish postal service provider) and flight delay prediction for the country’s national airline, Finnair.

About Author

Brian Dixon is a business and industry journalist with more than 20 years' experience writing about ports and logistics. A member of the Chartered Institute of Journalists, he has covered stories on six continents. He divides his time between the UK and East Asia