IBM and marine research organization Promare announced last week the first sea trials of the artificial intelligence (AI) that functions as the captain of the ship Mayflower, a step forward in building the autonomous ship.
The new Mayflower will cross the Atlantic unmanned, and controlled by a non-human captain later this year. The test, which will take place on a manned research vessel off the coast of Plymouth in the UK, will assess how the ship uses onboard cameras, AI and edge computing systems to safely navigate around ships, buoys and other. obstacles that you are expected to encounter in the ocean when you go out to sea on September 16, 2020.
The autonomous ship will sail the original 1620 Mayflower route to commemorate the 400th anniversary of the famous voyage. Sailing from Plymouth, UK, to Plymouth, Massachusetts, without a human captain or crew, she will likely become the first full-size, fully autonomous vessel to cross the Atlantic. In this way, the mission will foster the development of autonomous commercial vessels and transform the future of marine research.
The market for autonomous boats will grow from $ 90 billion today to more than $ 130 billion by 2030. Many of those around today are robots that do not dynamically adapt to new situations and rely heavily on control of operators. Using an integrated set of cutting-edge technologies such as AI and the Mayflower cloud you will have complete autonomy.
The autonomous ship will use IBM artificial intelligence and edge computing systems to make decisions at sea , even without human intervention. Although construction of the trimaran is nearing its final phase in Gdansk, Poland, an AI-based prototype of the ‘captain’ will float on the manned Plymouth Quest, a research vessel from the Plymouth Marine Laboratory in the UK. The March tests will take place in the waters of Smart Sound Plymouth, with the help of the human crew of the Plymouth Quest. Together, they will help determine how the autonomous captain performs in real-world maritime scenarios and provide relevant information to further refine the ship’s learning models.
The March maritime trials will run for approximately three months with a human captain at the helm. Then, in May, new tests will begin to assess the full range of the new Mayflower ship.
For the past two years, the Mayflower team has been training the ship’s AI models using more than a million nautical images collected by cameras in Plymouth Sound Bay (UK), as well as open source databases. The team also used an IBM POWER AC922 for the machine learning process, the same IBM POWER technology behind the world’s smartest AI supercomputers. Taken together and using IBM PowerAI Vision computer vision technology, the Mayflower captain should be able to independently detect and classify ships, buoys and other obstacles such as land, breakwaters and debris.
As the Mayflower will not have access to high-bandwidth connectivity during its transatlantic voyage, it will use a fully autonomous edge computing system running on Red Hat Enterprise Linux and IBM’s edge computing solutions, powered by multiple Xavier devices. NVIDIA on board. While at sea, the Mayflower will process data locally, increasing the speed of decision-making and reducing the amount of data flow and storage on the vessel.
In addition to meeting the overall mission objectives to reach Plymouth, Massachusetts, in the shortest time possible, the autonomous captain will use the IBM (Operational Decision Manager – ODM) rules management system to follow the International Regulations for the Prevention of Collisions at Sea (COLREG), as well as the recommendations of the International Convention for the Safety of Human Life at Sea (SOLAS).
The AI-based captain of the new Mayflower will use forecast data from IBM’s The Weather Company to help make navigation decisions. With these and other critical data such as depth and condition of the ship in mind, the autonomous captain is designed to operate independently in some of the most challenging circumstances. The ODM system also provides a completely transparent record of your decision-making process , avoiding “black box” scenarios . Faced with adverse scenarios, the autonomous captain of the Mayflower is also capable of analyzing without satellite connection what he is facing and deciding what measures to take through his technology.