Ulysses.

A Swarm of Robots for the Part of Earth We Forget

The ocean is 71% of the planet's surface, and the standard way to do anything useful in it is roughly this: rent a ship, which is expensive, staff it with people, who are also expensive, and send them somewhere for a while, which - you guessed it - is expensive. A single day of a research vessel can cost more than a used car. The result is that the ocean, despite being enormous and mostly the reason the planet is habitable, is monitored about as carefully as the back of your fridge.

Ulysses is a bet that this is a pricing problem, not a physics problem. Founded in 2023 and headquartered in San Francisco, the company builds autonomous underwater and surface robots that are cheap enough to deploy in swarms - the idea being that ten robots that each cost $50,000 can do things that one $5-million ship cannot, chiefly because you can leave them out there. The company's public tagline is that it keeps oceans "safe, healthy, and prosperous," which is the kind of phrase that means nothing until you notice that all three of those words are also things navies, energy companies, and conservation groups will pay for.

The origin story is almost annoyingly charming. In 2023, co-founder Jamie Wedderburn was on a surf trip to the west coast of Scotland and learned that volunteers were hand-planting seagrass seeds - in the North Atlantic, in bad weather, by diving. Seagrass, it turns out, is a quietly heroic plant: it covers about 0.1% of the ocean floor, stores carbon faster than a rainforest, filters water, and shelters fish. It is also disappearing at roughly 7% a year. The market response to this had been, essentially, cold volunteers. Wedderburn and his co-founders - Akhil Voorakkara, Colm O'Brien, and Will O'Brien - were robotics engineers with, by their own admission, no marine-biology credentials whatsoever. So they did the thing that separates a company from a hobby: they called people who did have the credentials, listened, and then built a robot to do the diving.

The trick is the payload, not the robot

Here is the part that makes Ulysses interesting as a business rather than as a nature documentary. The flagship robot, called Mako, is modular. It's about two meters long, weighs around 400 pounds, dives to 5,000 feet, and runs for 72 hours at a stretch. But the useful bit is that its payloads swap in and out. The same Mako that plants seagrass off Australia one week can, the next week, inspect a submarine power cable in the Baltic Sea. One is a climate-restoration job paid for by a conservation group; the other is a critical-infrastructure job paid for by an energy company or a government. The robot doesn't care. It's the same hull.

This is a genuinely clever piece of business design, because it means Ulysses isn't picking between the small, feel-good market (restoration) and the large, less-photogenic market (defense and offshore energy). It's selling one platform into all of them and letting the payload decide which invoice gets sent. A cynic would call this hedging. A less cynical person would call it the reason the same product can be funded by both Lowercarbon Capital, a climate fund, and Booz Allen Ventures, whose interests are somewhat more Pentagon-adjacent.

Cheap is the whole strategy

Most autonomous underwater vehicles are built for the deep, dark, crushing abyss, which is impressive and mostly irrelevant to the work that actually needs doing. Ulysses inverts this. Its pitch is that the valuable jobs - restoration, cable and turbine inspection, coastal awareness - happen in relatively shallow water near coasts, so building a vehicle that survives 20,000 feet is over-engineering you then have to charge customers for. By aiming at the coasts and driving the price down, Ulysses claims restoration roughly 100x faster than volunteer diving and inspection at a fraction of vessel cost. Whether those exact multiples hold across every deployment is the sort of thing that's hard to verify from the outside, so treat them as the company's figures rather than gospel. The direction, though, is clearly down-and-to-the-right on price.

The commercial proof arrived early. Ulysses booked close to $1 million in revenue in its first year, selling not robots so much as outcomes - a kilometer of restored seafloor, an inspected asset - to governments in Florida and Australia and to conservation groups. Then, in April 2026, Andreessen Horowitz's American Dynamism fund led a $38 million Series A, bringing total funding to about $46 million, with Booz Allen Ventures, Lowercarbon Capital, ReGen Ventures, Superorganism, and Harpoon Ventures along for the ride. The money is earmarked for scaling Mako and pushing two more machines - the Kraken recharge system and the Leviathan surface vessel - from prototype toward production.

What Ulysses is really building, if it works, is persistence: robots that live at sea, recharge themselves via Kraken, and quietly keep an eye on things without a ship or a crew. That's a modest-sounding ambition with an enormous surface area, which happens to describe most of the planet.