Deep-sea robots mostly work day shifts. A ship sails out, lowers one on a cable, films the bottom for a few hours and winches it back before dinner. That schedule is great for finding new animals and hopeless for catching anything that moves at the speed of a season.<\/p>\n
MBARI left one down there.<\/p>\n
The Benthic Rover II crawls the abyssal plain at Station M, a research site 140 miles (225 kilometers) off the central California coast where the water runs 13,100 feet (4,000 meters) deep. It drives 32 feet (10 meters), stops, pushes two clear chambers into the mud and waits 48 hours while sensors count the oxygen vanishing inside them. Every worm, microbe and sea cucumber in that patch of sediment is breathing. Falling oxygen tells the researchers how much carbon the bottom is eating, and that tells them how much is staying down there instead of coming back up as carbon dioxide.<\/p>\n
NOAA handed that measurement a new audience this month. On August 19 the agency published a mining company’s application to work the deep Pacific floor for battery metals and opened a public comment window that closes on October 19.<\/p>\n
Benthic Rover II measures 8.5 feet long, 5.6 feet wide and 4.9 feet high (2.6 by 1.7 by 1.5 meters), which MBARI<\/a> puts at about the size of a small car. It rides on a pair of wide rubber tracks that spread its weight so it does not sink into sediment with the consistency of pudding. Titanium, plastic and pressure-resistant syntactic foam make up the body, since steel and seawater have a well-documented disagreement. The hull is rated to 19,700 feet (6,000 meters), which is deeper than Station M by half again.<\/p>\n The electronics average two watts across a full year. Batteries have to last twelve months in the dark with nobody aboard, so the computer has to sip.<\/p>\n Paul McGill, an MBARI electrical engineer, described the software problem in 2021: “nobody is there to press a reset button.”<\/p>\n A ship recovers the rover roughly once a year. Engineers download the data, swap the battery and put it back over the side, where it free-falls for about two hours to reach the seafloor. To come home it drops a 250-pound (113-kilogram) ballast weight and floats up on its own. Holding position for two days at a time is the hard part of any subsea job, which is why a commercial inspection drone that can hang in front of one weld for 24 hours<\/a> counts as a selling point.<\/p>\n Between those visits nobody can call it. MBARI sends a second robot instead. A Wave Glider sails out from shore, pings the rover acoustically from the surface, collects its status and sample data and relays everything to Moss Landing by satellite. The institute’s own summary of the arrangement is that it sends a robot to find its robot.<\/p>\n MBARI reported seven years of continuous operation in November 2021, with the rover spending 99 percent of its life on the seafloor. Counting from the first Benthic Rover deployment at Station M in 2011, the machines had covered more than 6.8 miles (11 kilometers) of abyssal plain by October 2018. That is 6.8 miles in seven years.<\/p>\n Between November 2015 and November 2020 the rover logged a substantial increase in the dead phytoplankton and plant-rich debris landing on the seafloor. Dissolved oxygen in the water just above the bottom fell over the same stretch. Both results came from one machine, at one site, across the same five years.<\/p>\n The abyss turned out to be a lot twitchier than the textbooks had it. MBARI has reported pulses of marine snow that deliver in a few weeks as much food as decades of ordinary sinking debris would. A camera that visits once a year misses that completely, which is the whole argument for leaving a machine on the bottom instead of renting a ship for a week. Christine Huffard, a senior research specialist at MBARI, works on exactly that question of what gets stored in abyssal sediment and what gets eaten. The Station M time series<\/a> she works from started in 1989 and sits at 34\u00b050’N, 123\u00b000’W.<\/p>\n For years the R\/V Western Flyer handled recovery duty. That ship made its final MBARI voyage in November 2022, and the destination was Station M. MBARI transferred the vessel to the Florida Institute of Oceanography on November 15, 2022, and it sailed for St. Petersburg through the Panama Canal that January.<\/p>\n The replacement took a while. MBARI welcomed the R\/V David Packard, 164 feet (50 meters) long and 62 percent larger than the Western Flyer, into the fleet in 2025.<\/p>\n The Metals Company USA filed an amended consolidated application with NOAA on March 19, 2026, asking for an exploration license and a commercial recovery permit in the Clarion-Clipperton Zone, the nodule field in the Pacific southeast of Hawaii. NOAA found the filing compliant and published notice of it in the Federal Register<\/a> on August 19, 2026. Written comments close on October 19. A virtual hearing runs on October 13 from 3 to 5 p.m. ET, capped at 1,000 attendees and three minutes a speaker.<\/p>\n The nodules are the reason. Manganese, nickel and cobalt sit inside those lumps in roughly the proportions a battery plant wants, which is why a collector machine that works like an industrial vacuum<\/a> is being built to pull them off the bottom by the million tons. Japan has a state agency building its own deep-sea drone<\/a> to scout rare earths at 19,700 feet (6,000 meters).<\/p>\n Station M is not in that zone. Nobody is proposing to mine it. Both patches of seabed lie at about 13,100 feet on soft abyssal mud, and only one of them has a record of carbon supply and demand running back to 1989.<\/p>\n British researchers went back in 2023 to a patch of the Clarion-Clipperton Zone that a test collector drove across in 1979. The tracks were still there 44 years later, with a 26-foot (8-meter) strip of seabed cleared of nodules and two furrows where the machine passed. Daniel Jones of the UK’s National Oceanography Centre led the expedition and published the results in Nature<\/a> on March 26, 2025. Sediment macrofauna, mobile deposit feeders and even some large sessile animals had begun to re-establish themselves. The communities inside the disturbed tracks remained altered anyway.<\/p>\n That is one 26-foot stripe from a single test run in the 1970s, revisited by a ship that had to go looking for it. The rover at Station M has been taking the same class of measurement in the same place every 48 hours, which is the difference between a photograph and a security camera.<\/p>\n MBARI made the mining argument itself back in 2021, and it was less dramatic than the one you hear from either side of the permit fight: with companies looking to pull minerals off the deep seafloor, a long baseline tells you what the bottom looked like before anybody touched it. MBARI calls Station M the only site where carbon supply and carbon demand have both been recorded as a time series, and one of three long-term abyssal time-series sites in the world. The other two do not sit anywhere near a nodule field either.<\/p>\n NOAA’s docket on the application, NOAA-NOS-2026-1751, had collected 51 public comments as of August 30, 2026. The agency published a notice of intent to prepare an environmental impact statement on August 19, the same day it published the application itself.<\/p>\n","protected":false},"excerpt":{"rendered":" Deep-sea robots mostly work day shifts. A ship sails out, lowers one on a cable, films the bottom for a … <\/p>\nOne service call a year, and 99 percent of its life on the bottom<\/h2>\n
Five years of readings turned up more food and less oxygen<\/h2>\n
The ship that used to pick it up now works out of Florida<\/h2>\n
NOAA opened a comment window on August 19<\/h2>\n
A 1979 mining track still shows up on camera<\/h2>\n