Designed by not-for-profit research company ProMare, the Mayflower Autonomous Ship (MAS) will navigate the 5,182 kilometers (3,220 miles) between Plymouth, England and the US city of Plymouth in Massachusetts. Helming the vessel will be IBM’s AI, featuring a menagerie of machine learning, edge computing, and cloud technologies. According to Brett Phaneuf, the managing director of Mayflower project, the idea for the vessel came a few years ago after the Plymouth Council suggested designing and launching a replica. He writes: MAS will carry research pods full of sensors that scientists will use “to conduct persistent, ground-breaking research in meteorology, oceanography, climatology, biology, marine pollution and conservation, and autonomous navigation.” But the real draw here is that it’ll make the crossing almost entirely on its own. IBM’s computer vision, LIDAR, and RADAR systems will ensure the ship doesn’t collide with other objects at sea, handles rough weather well, and stays on course. It’s equipped with onboard edge-technology so that it can continue on its mission even if it loses contact with IBM and ProMare. The original Mayflower utilized cutting-edge navigation and power technology for its time, namely the stars and wind. While MAS certainly won’t be charting its voyage using a compass and sextant, it will use a similar form of power: natural sunlight and wind. Solar panels and a single wingsail will provide the energy needed to power the vessel’s motors and charge its batteries for night use. According to ProMare, the vessel will travel at about 20 nautical miles (knots) per hour – nearly ten times faster than the original Mayflower’s plodding 2.4 knots per hour. The MAS project is designed to test out AI technologies at sea and to inspire future researchers to brave the Earth’s oceans, 95% of which remain unexplored. Phaneuf writes: MAS launches in September of 2020. Once underway, ProMare says it’ll make the voyage accessible to the public in virtual and augmented reality thanks to support from the University of Birmingham.
