Inspired by a voyage from centuries ago, marine researchers and climate scientists teamed with tech titans like IBM to showcase how crewless vessels can help protect and preserve oceans. The goal of the US$1.3 million Mayflower Autonomous Ship (MAS) project? Spark a new era of ocean research that will help scientists analyze information more efficiently.

“To understand our oceans of the future, we really need to collect as much data as possible—and that’s something that autonomous vessels will really allow us to do,” says Rosie Lickorish, software engineer, IBM, Southampton, United Kingdom.

More than 400 years after pilgrims crossed the Atlantic in the original masted Mayflower to establish a U.S. colony, the MAS team aimed to retrace the route. But this time around, the trip was made in a small vessel powered largely by solar panels and a wind turbine—sans any human crewmembers but loaded with AI and machine learning capabilities.

To create the vessel, nonprofit marine research organization ProMarecollaborated with the University of Plymouth, and IBM providing tech reinforcements, like servers, AI, cloud services and edge computing. The team packed the ship with three research pods and an array of sensors and other gadgets so it could probe issues such as maritime cybersecurity, marine mammals, sea-level mapping and the impact of ocean plastics.

The project also aligns with the United Nations’ Decade of Ocean Science for Sustainable Development, which aims to reverse a decline in ocean health. Along with capturing and analyzing data on things like the ocean’s temperature, salinity, oxygen, pH and pollutants in the water, the team collected underwater audio streams and use video to study the large-scale currents and surface flows of the ocean.

“We wanted to build something that could create information—using edge computing—from vast amounts of data that it gets, send that information back and drive the cost down,” says Brett Phaneuf, project director and ProMare co-founder. “That way we could sort of democratize ocean and climate science, because we can give these tools away for free. And then anybody can build a ship.”

Knowing that threats like storms and connectivity issues could impact any project voyage, Phaneuf made sure the team integrated risk management from the start. “When you deal with the ocean, everything you put in it or on it has a very high likelihood it’s not coming back,” he says.

Project partners identified as many scenarios as possible and then iterated their way to a vessel that was up to the job. For example, Marine AI handled the  design of the software in partnership with IBM and with support from NVIDIA, developing a virtual navigator called A.I. Captain. MSubs used aluminum and composite materials to create a ship that’s lightweight but  can stand up to the ocean’s punishing waves. And Iridium’s satellite communications system helps the vessel stay hyperconnected so it can generate nonstop, real-time data.

The ship’s first mission last year was aborted after three days because of mechanical problems. Unbowed, the team completed several sea trials before achieving its first transatlantic crossing on June 5—a 40-day trip that spanned approximately 3,500 miles (5,633 kilometers).

Through it all, the team’s agility and innovative thinking overcome any waves of disruption. Most notably, when technical problems resurfaced near the end of the ship’s voyage, project leaders changed course. Instead of docking the ship at the planned U.S. port in Massachusetts, the team diverted it to Halifax, Nova Scotia, Canada.

Yet the successful crossing was just the start—the team plans to have the ship explore other oceans and coasts as it conducts climate research.

“There’s still things to be discovered,” Phaneuf says. “If we have a zero-risk philosophy as the criteria for doing new things, we are doomed.”

Image credit: IBM