A ship piloted by an “AI Captain” will be taking to the seas this month, according to PR Newswire. Promare, a marine research organization, announced that the Mayflower Autonomous Ship (MAS) will be able to self-navigate across the Atlantic later this year.
- A ship piloted by an AI Captain will cross the Atlantic later this year.
- The technology is set for manned trials off the coast of Plymouth, UK.
- The Mayflower Autonomous Ship (MAS) will take the same route the original Mayflower followed 400 years ago.
Before the MAS makes that voyage, a trial of the technology is planned for a manned research vessel off the coast of Plymouth, United Kingdom. The trial on the Plymouth Quest will allow for evaluation of how the AI Captain uses cameras and various computing systems to navigate.
The name might be the same, but the Mayflower Autonomous Ship looks nothing like its namesake. (Photo: Promare)
Taking the AI Captain for a test run off the coast of the United Kingdom
These sea trials are due to last for two months with the ship’s human crew manning the helm. Initial testing will determine how well the AI Captain receives input from the ship’s radar, AIS, GPS, and navigation. It will also ensure that visibility data is accurately processed. Later testing will look at cameras, computer vision, and autonomy capabilities.
Much like an autonomous car, an autonomous ship must be able to navigate a variety of challenging situations. These include other ships, buoys, rocks, and ocean hazards that could impede its path. One big difference with this ship is that it won’t simply be able to pull over if it runs into trouble in the middle of the Atlantic.
The MAS will take the same path as its namesake, the 1620 Mayflower, and sail from Plymouth, UK to Plymouth, Massachusetts to commemorate the 400th anniversary of the original voyage. There will be no captain and no crew on the ship. This will make it one of the first full-size, fully autonomous ships to make a transatlantic voyage.
If you’re thinking it will be a big disappointment to not have the chance to dine at the captain’s table on your next luxury passenger cruise, fear not. Although this technology has multiple applications, it is currently targeting research vessels and cargo transport, which is expected to grow at roughly 4.4 percent in the next years, according to Allied Market Research.
It’s taken two years for the Mayflower’s team to reach this point. The ship’s AI was trained using over one million nautical images from cameras in Plymouth Sound along with open source databases. It required an incredible amount of processing power, so the team used the same technologies that are used in modern AI supercomputers.
Ready for the unexpected in the middle of the ocean
It should allow the AI Captain to accurately identify objects in the water. This includes ships and buoys as well as land, breakwaters, or floating debris. What a ship encounters on the ocean is unpredictable and the AI Captain must be prepared for any situation. The unpredictability of ocean travel is a challenge even for humans, with anywhere from 75 to 96 percent of marine accidents involving human error, according to Allianz.
The Mayflower won’t have access to high-bandwidth connectivity, so everything it needs to process information is right on the ship. Being able to process data locally increases the speed of its decision-making process and reduces the amount of data flow and storage on board the ship. When a connection resumes, it can confirm its course with humans on shore as a safety check.
One of the biggest challenges for Mayflower will be the weather. To help navigate rough weather, it will use data from The Weather Company to determine the best possible route. It will even be able to navigate around the unexpected, like a cargo ship that loses part of its load necessitating a new path for the MAS.
Boris Johnson, then Foreign Minister, now Prime Minister of the UK, stands next to Boaty McBoatface, an autonomous submarine exploration vehicle. Such small-scale developments of aquatic AVs will lead the way for full-scale implementation. (Photo: Getty Images)
WHY THIS MATTERS
Eliminating the need for a captain and crew on cargo ships makes more room for cargo once there’s no reason to have a kitchen or sleeping quarters. It also reduces the chance for human errors, which can be costly to companies. Incorporating artificial intelligence into cargo ships makes companies more profitable and the ocean safer.