He spent twenty years pulling wrecked ships off reefs. Then he sat down in a shared Cambridge office and started writing software so the wrecks would stop happening.
The idea for Sea Machines Robotics came to Michael G. Johnson on an oil-spill response job in the Arctic, which is a sentence that sounds fictional but is a matter of public record.
Johnson was a Vice President at Crowley Maritime and its affiliate TITAN Salvage. Titan was the outfit that specialized in unwrecking ships - the Costa Concordia, that sort of job. He had spent two decades in offshore oil and gas, heavy-lift ocean transport, and marine salvage, with postings that read like a Melville itinerary: Singapore, the Arctic, Europe, Australia. He had watched, up close, what happens when the person on watch on a very large piece of steel makes a small, wrong decision at three in the morning.
The Arctic spill response, by his own telling, was the moment the idea crystallized. Coordinating vessels in cold, dull, dangerous water was, he decided, a job humans should not have to do. In January 2015 he moved to Boston, took a desk in a shared Cambridge office, and started Sea Machines Robotics with one employee: himself.
The first demonstration platform was a 25-foot steel twin-screw azimuth German tug, docked at Boston Harbor Shipyard. This is a very specific boat and it is worth noticing - the founder did not start on a whiteboard. He started on a tug.
Sea Machines' pitch, then and now, is not that ships should be crewless. It is that most of what happens on a working vessel is routine: hold this heading, watch for this obstacle, run this survey grid, dump this cargo here, come back. Johnson likes to call these the "three Ds": dirty, dull, dangerous. Humans, he argues, are good at a lot of things. Continuous vigilance across long, repetitive missions is not one of them.
So Sea Machines builds two things. The first is autonomous control - the software that lets a workboat run a mission with a person supervising from shore instead of standing at the helm. The second is a computer vision perception system, called AI-ris, that acts as a navigational sensor: it looks at the water and tells the vessel what is out there, in a way that no radar or AIS transponder can. In 2018 that perception system was trialed on an A.P. Moller-Maersk container ship, which is the maritime industry equivalent of getting your first pilot to fly a 747.
In 2021 the company staged something called The Machine Odyssey. A vessel, autonomously navigated, covered more than 1,000 nautical miles over 13 days, commanded from a control room in Boston. This is the sort of thing you do when your board asks whether the technology actually works and you would like to answer with a photograph rather than a slide.
Cap table includes Accomplice, Toyota, Brunswick, Huntington Ingalls Industries, Launch Capital, LDV Capital, Geekdom Fund, Techstars.
In November 2024, Sea Machines announced that David "Chip" Wasson - a former Navy flight officer, investment banker, and Huntington Ingalls Industries executive - would take the CEO seat. Johnson would move to President and Chief Technology Officer.
Founder-to-CTO transitions are usually one of two stories. Either the board pushed the founder out, or the founder decided that scaling a company past a certain point requires a different job than the one they wanted to be doing. Johnson's version, on the public record, is the second one. Huntington Ingalls, which builds nuclear submarines and destroyers, had been on Sea Machines' cap table since 2020. Wasson's arrival was a signal: defense is the next chapter, and it will be sold in a language Sea Machines' engineer-founder does not natively speak.
What Johnson gets back, in exchange, is time to be the engineer. The company still ships product - SM200, the wireless remote helm system that earned USCG, ABS, and later Bureau Veritas type approval; SM300, the autonomous command system for workboats; JETsense, a voyage-control partnership with HamiltonJet; AI-ris, the vision sensor. Each of these has to be maintained, extended, and integrated into the messy variety of hulls, engines, and radars that make up the commercial fleet.
"This is the right time for Chip to step aboard, take the helm, and duly navigate our company," Johnson said in the announcement. It is a nautical sentence written by a person who cannot help writing nautical sentences.
"We build new software deployed on proven off-the-shelf hardware," Johnson has said. It is an unglamorous position that has kept the company alive. Sea Machines sells autonomy that runs on the industrial computers, radars, and thrusters that shipyards already trust.
Johnson has argued in interviews that commercial markets are healthier competitive spaces than defense sectors for early autonomy work. The 2024 CEO handoff to a defense industry veteran suggests the company is now old enough to change its mind.
Phase one: build autonomy that matches a skilled ship captain. Phase two: build autonomy that surpasses even the most experienced ones. He has been public about phase one being nearly done.
"We are not only developing cutting-edge autonomous technology, we are already commercializing it as usable products."
"An autonomous system that's always tracking brings predictability to operations and reduces human error-caused incidents."
"This is the right time for Chip to step aboard, take the helm, and duly navigate our company."
Founder, president and chief technology officer of Sea Machines Robotics, a Boston company building autonomous control and perception systems for commercial marine vessels.
No. In November 2024 David "Chip" Wasson became CEO. Johnson moved to president and CTO.
Texas A&M University, where he earned a degree in marine engineering.
He was a vice president at Crowley Maritime and its affiliate TITAN Salvage, and spent two decades in offshore oil and gas, heavy-lift ocean transport, and marine salvage.
About $44 million in total, with backers including Accomplice, Toyota, Brunswick, and Huntington Ingalls Industries.