Automation on the High Seas: AI-driven Cargo Ships Are Here to Stay

With the world’s eyes set on automobile manufacturers like Elon Musk’s Tesla, it is easy to lose track of automation in other areas. Automated driving is set to revolutionize the way we move around within the next decade. This is certainly breaking news. The new form of transport promises to cut down traffic accidents and the number of fatalities on the road. It will get us from A to B in a safer, and in all likelihood, more timely manner.

But not every sector or type of transport gets the same media attention. Many changes go by quietly, even though their impact is likely to be felt across the board. This is the case of automated cargo shipping, which promises to radically change the way we connect or trade with the rest of the world. Ever since the invention of the cargo container, trade links are defined by deep-sea port networks and intercontinental maritime routes.

Automated shipping is probably one of the most prominent examples of AI technology put into practice. Trials to eliminate the need for a captain, or crew, have been well underway for the last five years. But, until now, no company dared to fully explore or implement its existing potential, which perhaps comes to down to strict safety measures and backlash from trade unions. In other words, its commercial or operational viability is still under question.

IBM decided to break the silence. The IT giant recently announced that it will launch its flagship project, Mayflower Autonomous Ship (MAS) later this year, in September. The AI-equipped ship will “self-navigate” across the Atlantic, sailing from Plymouth, UK to Plymouth, Massachusetts in the US, recreating the original voyage, made by the Pilgrim Fathers in 1620. 

IBM is set to make waves with AI Captain

IBM’s venture is not purely symbolic or a nice way of commemorating a historical event. MAS is fully equipped with cutting-edge technology, which has not been applied anywhere else before. The objective is actually to see if the ship is able to withstand some of the most challenging circumstances on the planet, with minimal human intervention. Although IBM has kept a low profile until now, the project has been in the development phase for more than two years.

In the meantime, it has put together an international team, working in laboratories and research centers in Great Britain, Poland, and the US. The cherry on the cake is the AI Captain operating system, which comes with a number of state-of-the-art features, uniquely developed by IBM and its partner companies, such as NVIDIA. In fact, the team is comprised of experts from almost every single field we can imagine.

One of the weirdest plans was to introduce engineers from the Bitcoin Trading app, which is also an AI operating system. Due to the success that the software had seen in the finance industry and the rate at which it learned the curves and tendencies of the market, it was believed that it could be a good fit to navigational engineering. But, in the end, it was still decided to go with a completely new system from scratch.

What does this system contain? An Automatic Identification System (AIS), informed by over a million ultrasound nautical images. It also draws on IBM’s Operational Decision Manager (ODM), to develop hypothetical scenarios.

What does it mean in practice? The AI Captain is constantly re-evaluating the situation at sea, being hooked to numerous sensors, radars or navigation systems. While it will take some time before it reaches commercial shipping, the AI Captain is fully compliant with International Regulations for Preventing Collisions at Sea (COLREGs) or International Convention for the Safety of Life at Sea (SOLAS).

While the automated Mayflower is just a trial, IBM pointed to a recent study by Allied Market Research, which shows that the autonomous shipping market is projected to increase to almost $135 billion, by the end of the decade. There is a real prospect for growth and the technology is largely here already. Will we see fully automated ships before self-driving cars? Only time will show.

What is the economic implication?

AI always has something to do with economics, especially when it is trying to tackle such a large industry. The whole world floats (pun intended) on cargo ships and large freighters pretty much keeping the global economy alive.

Many would say that this industry does not require any technological advancement considering just how much it is able to deal with right now, however AI has one very specific advantage that it could bring into this sector.

Let's start with the obvious one. The salaries of the crew. Usually, a cargo ship would contain a couple dozen people operating it, not including the captain and the various officers helping him out. On average, a cargo ship crew would have around 30 people. According to data, the average salary of a cargo ship employee is around $40,000 per year. So, let's assume that there is 1 captain and 4 officers, the rest are regular workers. $40,000 times 25 that's already $1 million. Then the captain's salary which is around $80,000 per year, and the officers get somewhere around $55,000. In total, the cost of having a 30-man crew on a cargo ship is $1.3 million per year. That's not too much for one ship, but remember that most shipping companies have dozens of freighters.

AI has the capacity of bringing that 30-man crew to just 3 or 5 thanks to most of the maintenance being undertaken by the software. This could drastically decrease the costs for shipping companies, thus allowing them to offer much smaller prices to manufacturers that are willing to export their product.

Since the prices on shipping will decrease, it's safe to assume that the prices on goods themselves will decrease. This means that even though AI will take some jobs away in the shipping industry, it will generally open up a more affordable global market for the majority of the world population.

Additional innovations after AI

2020 tech innovation will most likely revolve around two sectors, AI and medicine. With the COVID-19 pandemic growing as we speak more and more innovators are trying to come up with creative solutions every single day. Which is an implication of what desperate times can do to human ingenuity.

When it comes to the cargo ship industry though, we've already seen AI being considered as a viable option, but there's more to come. Innovation always demands more, so we could start seeing shipping companies looking for additional ways to cut costs, and the only thing they can do this with is fuel.

That's right, there's a possibility that one of the largest contributors of global pollution may soon completely switch over to electricity rather than oil or gas. There's even discussions if there could be a mass production of nuclear-powered cargo ships. But that's a debate for the future.

What we need to focus on now is getting AI into this industry and having more affordable goods travel faster and more efficiently all over the world.