Department of Marine Science
Ulithi ROV Project
In October of 2012, shortly after presenting their results from EEL's C-DEBI Catalina ROV project at a scientific meeting, three EELs (Director Steve Moore and undergraduates Josh Ambrose and James McClure) were invited to join the Ulithi Marine Conservation Project. This project, under the direction of Nicole Crane, is a collaboration among a dedicated group of marine ecologists, the Oceanic Society, and the people of Ulithi Atoll, who are all working together on strategies for effective management of the coral reefs and fish populations upon which the people of Ulithi depend. Our unique contribution to this collaborative effort was to develop an ROV that could extend the accessible depth range beyond where the scientists had previously been able to collect data.
Over the next several months, Josh, James, and Steve, worked in the EEL at a feverish pace to develop a workable prototype of an ROV capable of being transported by air to Ulithi and, once there, diving to a depth of 150 meters to record high-defintion videos of that previously unseen deep-reef world.
In late June of 2013, all their hard work paid off: Josh, James, and Steve flew to Ulithi with their “Ulithi ROV.” They spend a lot of time in the air over a vast expanse of water and got to visit plenty of airports (some bigger than others): Monterey, LAX, Honolulu, Guam, Koror, Yap, and finally Ulithi. They arrived on the Ulithi island of Falalop on 1 July 2013, where they spent 12 days meeting the local community and their fellow scientists, visiting a few of the other islands around the atoll, scrambling to finish some not-quite-finished parts of the Ulithi ROV, and finally doing a few dives with it to begin the work of understanding Ulithi’s deep reefs and their connection to Ulithi's fish populations.
About Ulithi Atoll
Like other atolls, Ulithi is a ring-shaped coral reef that is thought to have grown upward from the former shoreline of a sinking volcanic island -- an island which has long since disappeared beneath the oceans' surface. Ulithi is located in the tropical Western Pacific Ocean, roughly 900 miles east of the Philippines. Its reef encloses a lagoon that's over 200 square miles in area, making it one of the largest atolls in the world. Inside the lagoon, the water averages 80-120 feet deep, but on the outside edge of the atoll, the reef drops precipitously to depths of over a thousand feet.
Although the original volcano is long gone, there are islands here. But they are not big ones. They are tiny “islets,” little more than raised sand bars on the reef. Most support a small forest of coconut palms and other tropical vegetation. There are roughly 40 of these islets scattered around the permeter of the atoll, but only four are inhabited. The largest of these, called Falalop, is about 1/2 mile in diameter -- much smaller than the CSUMB campus!
Politically, the islands of Ulithi are part of Yap State within the Federated States of Micronesia. Each island has one or more chiefs. One of the Ulithi islands, called Mogmog, is home to the High Chief of the region. He is the cultural leader for all the Ulithi islands, plus many other islands farther to the east.
Ulithi Atoll is among the most remote locations on the planet. There is no cell phone service and no internet. A supply boat comes out to the atoll only twice a year, bringing crates of food, water, fuel and other provisions. Everything else comes and goes via a small plane operated by a missionary organization.
Because of the islands' isolation, many of the traditional ways of life persist. Unfortunately, some customs and many valuable pieces of traditional knowledge were lost following WWII, when the islands and lagoon were temporarily converted to a massive, secret US Naval base during the war in the Pacific. At one point during the war, over 700 US Navy ships were anchored in the Ulithi lagoon. At that time, it was the largest naval facility in the world. As a consequence of US involvement in the region then and since, most people on the islands now speak English in addition to their native Ulithian language.
Archaeological and geological evidence suggests these islands have been inhabited for perhaps 1,500 years. We know the people must have lived sustainably off the land and sea for most of that time, or their descendents wouldn't still be here today. Unfortunately, sustainability has become elusive, even though the current human population of the islands doesn’t seem to be much larger than in the past. Some of the causes probably relate to physical changes in the islands, such as those caused by heavy construction during WWII. More recently, sea level rise has displaced fresh water in the sandy soil, making it harder to grow taro and other crops that were previously a staple part of the islander’s diet. Another factor is that many of the traditional methods for harvesting fish and other resources in a sustainable manner have been forgotten. Foods brought in to supplement reduced natural harvest are not as healthful as the traditional foods, so nutrition-related diseases, such as diabetes, are becoming more commonplace.
Understandably, the people of Ulithi are concerned about their future. We are excited to have joined this small, but dedicated team of researchers — invited by the people of Ulithi — to work together with the island community in solving some of these challenging problems.
About the Ulithi ROV
The Ulithi ROV belongs to a class of remotely operated underwater vehicles sometimes nicknamed “flying eyeballs.” It is essentially a steerable underwater closed-circuit video camera that allows a pilot on the surface to see on TV monitors what the ROV's cameras are “seeing” under water and to “fly” the vehicle around by remote control to look at different things.
The ROV isn’t just one object, but a complete system of interacting objects. The Ulithi ROV system includes these four major components:
- The ROV itself goes under water and consists of a pair of GoPro Hero2 cameras in custom-machined aluminum pressure-proof housings mounted to a plastic frame along with four SeaBotix thrusters for propulsion. There are high-powered LED video lights to bring out true colors in the dim blue light found at great depths. They also allow the ROV to operate at night. A separate waterproof pressure canister houses the electronics, including aParallax Propeller microcontroller (essentially a tiny computer that serves as the ROV's brain), Pololu TReX motor controllers, assorted sensors, and other electrical components.
- The Pilot’s Control Console remains on the surface with the human pilot. It features two TV monitors (one for each of the ROV's cameras). It also has joysticks and buttons used to control the ROV's movements and some of its accessories, like the cameras and video lights.
- The Tether is basically a communication cable that carries the pilot's remote-control commands from the surface down to the ROV and relays live video from the ROV's cameras back to TV monitors sitting in front of the pilot on the surface. The tether is 170 meters (about 560 feet) long and is stored on a big spool with a hand crank to reel it back in.
- The battery box contains two 12-volt SLA batteries and is filled with vegetable oil for pressure compensation. These batteries supply power to the ROV through a 10 meter long cable. This gives the ROV some flexibility to move around independent of the heavy batteries, but is still close enough to receive most of the battery power without excessive power loss through long wires. The battery box dangles near the end of the tether, where it is attached 10 meters from the ROV.
More about EEL's ulithi visit in 2013
It’s a long way from Monterey to Ulithi.
And there’s a lot of water between here and there.
Flying over 7000 miles of open water really helps you develop a new appreciation for the enormity of the Pacific Ocean. It also leaves you in awe of the incredible skill of the original island navigators who used hand-carved canoes to travel between tiny specs of land scattered in the middle of this incredible vastness! Some of the finest traditional navigators in the world are known to have come from Micronesia.
Fortunately for us, planes travel much faster than canoes and have GPS navigation. Our particular route from Monterey passed through airports in Los Angeles, Hawaii, Guam, Palau, and Yap, before finally reaching Ulithi. There was no additional air fare to get off for a few days in the Republic of Palau, so we had planned a little side adventure and spent four days kayaking and snorkeling our way among Palau's beautiful Rock Islands and its unique coral reefs before continuing on our journey via Yap to Ulithi.
United Airlines could take us only as far as Yap Island in the Federated States of Micronesia, roughly 7,000 miles from CSUMB, but still 100 miles from Ulithi. The final leg to the island of Falalop on Ulithi Atoll required travel on a small propeller plane operated by Pacific Mission Aviation. Our pilot, Amos, was amazing. He managed to squeeze all of his passengers, their luggage, and all their science equipment (including 2 ROVs) into the tiny plane and make a picture-perfect landing on the short Ultithi airstrip, the two ends of which meet the ocean on opposite sides of Falalop island.
Day 1: We are greeted at the “airport” by Jon (“Junior”) Rulmal, Director of Ulithi’s Conservation Program and our host and guide. Junior is a native of Ulithi, but also lived on the US mainland for a while, so he serves as our cultural liason. He helps us understand local customs and interact respectfully with the local villagers on Falalop and the other Ulithi islands we visit. From the airport we walk all the way across the island (7 minutes) to our base of operations, the Ulithi Adventure Lodge, where we meet the other members of our visiting team, including several marine ecologists (Nicole, Giacomo, Avigdor, Michelle, and Peter) and a San Francisco based physician (Ricardo). After the meet and greet, we move ourselves and our ROV equipment into our room and set up the room as a mineature electronics workshop to complete a few remaining details on the ROV.
Day 2: We get a rare opportunity to visit a nearby uninhabited island (Gielap) where hundreds of sea turtles come ashore at night to nest and lay eggs. During the day, we snorkel on the coral reef there, watch zillions of hermit crabs scuttling underfoot, dine on a beach-cooked meal of coconut crab, fresh-caught fish, and breadfruit, and watch a gorgeous sunset while waiting quietly for the turtles to come ashore. About an hour after dark, Junior escorts us a green turtle that has begun to lay eggs in a sand pit it dug under a tree. The ride home is surreal. Three boats speeding invisibly across a dark sea. Stars glittering above against a jet black sky, bioluminscent plankton glittering like blue-green glowing embers in the spray kicked up by the boat.
Day 3: Josh and Steve spent the day working on the Ulithi ROV, while James took our Catalina ROV to the island of Asor to check out some of the shallow reefs and the old military wrecks there. Lawrence, one of our Ulithian boat drivers starts operating the ROV and masters it quickly!
Day 4: This day is a special treat and a great honor. We have been invited to the island of Mogmog, home of the high chief. Mogmog is the most traditional of all the Ulithi Islands. The ecologists will be meeting with the chief and villagers to share results from their research the previous year. We will be introducing the new research we’ll be doing for them with our ROV. Unfortunately, our Ulithi ROV isn’t ready yet, so we bring one of our smaller C-DEBI ROVs to pass around for “show and tell.” Junior is visibly nervous about this visit. Western culture and traditional Ulithian culture are different. He wants us to make a good impression. The three of us get dressed in “thus,” the traditional garment worn by men. A thu is a piece of long cloth wrapped between the legs and around the waist to form something resembling a hybrid between a skirt and a loincloth. It has to be tied and tightened just right, or it can fall off. Junior warned: “Don’t let your thu fall off, especially in front of the chief...that would be bad!”
Day 5: All three of us spent all day working on the ROV. At this point we’re getting pretty worried. Finishing up the few remaining steps to get the ROV operational is proving far more challenging and time consuming than we had hoped. When we got to Ulithi, we figured we’d have the Ulithi ROV in the water within a day or two. Now it’s day 5, and we're working late into the evening, and it’s still not functional, in fact, we’ve suffered some setbacks, because some things that were working earlier are no longer functioning properly. For example, modifications to the code needed to control the video lights seem to have rendered our thruster control non-functional. At least we got the LED lights sealed in epoxy, so they’ll be waterproof. That’s about the only concrete progress we can report today. Our tiny room in the lodge, which is serving as a dormitory for the 3 of us as well as a machine shop and electronics workshop, has become a complete disaster area. Word has gotten out, and other members of the team are beginning to stop by with their cameras to document the incredible degree of chaos. Given that we have to allow at least two days at the end of the trip to get everything rinsed, thorougly dried, disassembled, and packed, we've now passed the half-way point on our diveable days and don't have a working Ulithi ROV. We are glad we brought a Catalina ROV as a backup. It can't go deep, but at least it can do some shallow work and serve as our “show and tell” ROV.
Day 6: Another day devoted entirely to ROV construction. And again, everything is taking longer than anticipated. In the early morning, we were optimistic that we’d have it in the water by noon. By noon we were hoping for late afternoon. By late afternoon, we were hoping for a dive before nightfall. Finally, as the sun is about to set, we are ready. But there’s no boat available. Not a group that gives up after getting this close, we decide that a shallow test dive is better than no dive at all. Moreover, we notice that our tether might be just long enough to reach all the way from shore to a small canyon, about 30 feet deep, slicing through the reef crest near the lodge. In anticipation of having to work without a boat proper, we has brought along a bright orange kid's inflatable rubber boat: The Explorer 200! We used it to float the heavy battery box out to the edge of the canyon and launched the Ulithi ROV. To make a long story short, it worked! (Sort of.) We had a few minor issues, like a bug in the software, so it turned right when we told it to turn left, but hey, at least it didn't flood and sink to the bottom! And, as it turned out, doing our test drive during sunset enabled us to test how well the video lights worked in dim conditions. (The photo of the ROV with its lights on coral above was taken during this test dive.)
Day 7: This morning, in anticipation of our first deep dive later in the day, we got to work fixing some of the problems that had shown up during the test dive yesterday. We also discovered that our oil-filled battery box had leaked a lot of oil during the night, so we had to deal with that. (Fortunately the mess was contained within a large plastic tub.) We had run out of fresh vegetable oil to use, so we had to get some used oil from the kitchen. It had been used to fry fish for the previous night's dinner. Kinda stinky and gross, but it was either that or go without any deep ROV dives -- something unthinkable after so many months of hard work and anticipation. Today is Sunday, so we took a brief break from our work to attend the village church service and see what that was all about. It was a Catholic service delivered in Ulithian. People mostly sat on the concrete floor fanning themselves to stay cool. Altar decorations included an array of large artillery shells from WWII used as flower vases. That afternoon, we put the Ulithi ROV on a boat and headed out to a steep drop-off near Asor Island. Some other members of the team had seen it the previous day during a reef survey and recommended it to us as a good deep dive site. Unfortunately, the wind conditions were not cooperative, so we had to abandon that location. Instead, we went to the reef drop-off near the lodge. We had a pretty successful dive. We worked our way down the wall to a depth of about 40-45 meters (roughly 150 feet). Along the way we saw (and recorded) amazing life, including corals, crinoids, sponges, and a variety of fish. We even got a recording of a pair of Silvertip Sharks, a species that inhabits deep reef drop-offs, but is rarely seen in shallower water. Unfortunately, before we could go any deeper than 45 m, our forward camera started acting flaky — spontaneously turning itself on and off at odd intervals. With panicked visions of a leak destroying our ROV, we hastily pulled it up and brought it back to our room/workshop for a diagnosis. As it turns out, there was no leak. However, one of our leak detection circuits, which had been too-quickly slapped together in our rush to get in the water, came loose and shorted out against the inside of the metal pressure housing. That lowered the voltage to the camera control circuit board and caused the erratic camera behavior. Fortunately, it caused no serious or permanent damage.
Day 8: We had high hopes of a second deep dive today, but we’ve discovered more problems with the ROV circuitry and/or software. Voltages in several places aren't what they should be. While diagnosing and fixing those problems, we’re also trying to complete some unfinished features. For example, the ROV has a depth sensor, compass, and battery voltage sensor already built into the ROV hardware, but we have not yet finished writing the software to read and display the sensor values. Those “upgrade” attempts didn’t work very well, so we eventually gave up and focused on getting the essentials ready, to make sure we're ready for a dive tomorrow. The rest of the group headed out to some far away islands to survey reefs there, while we worked on the ROV. By evening, we thought we were close to having everything working again, but then one of the cameras started acting silly about 10:00 pm. Hope began to fade. Three hours later (1:00 am), we still had not resolved the problem. This certainly did not bode well for a deep dive the next day, but we were completely exhausted and figured we might make matters worse by working on the ROV while half-conscious, so we went to sleep.
Day 9: This morning all the boats are out taking other members of the team to present preliminary results to villagers on other islands, but one is supposed to come back for us this afternoon to take us out for what will likely be our last chance for a deep dive. We work frantically to get the ROV prepared. By 1:30 pm, we think it’s ready. At least it seems to be working. Fingers crossed. That afternoon, the winds are good, so we take Ulithi ROV to the Asor dropoff with the goal of diving to our maximum rated depth of 150 meters (500 feet). All systems check out, so we begin our descent a little after 3pm. Eighty-five meters down we encounter a vertical wall full of huge pink and yellow sea fans. At 95 meters, we get 4 full minutes of good video of a large Silvertip Shark circling the ROV. At 4:00 pm, we reach 100 meters (330 feet), and all systems are still working fine. Whew! Lots of colorful sponges, sea fans, and whip corals. Then, at about 130 meters depth, James reports a problem. One of the thrusters seems to be acting funny. Not good. Gradually over the next several minutes we lose control of more thrusters. Before long, the ROV is immobile in the water, drifting somewhere over 400 feet below us. Then we lose the video lights. Then we loose camera control. We're all thinking the same thing, but nobody wants to say it. Do we have a leak? Fortunately, we lost camera control while both cameras were still running, and we can see the live video images on the pilot's TV monitors, so it can't be a total disaster down there. Hoping the problem might be something simple on the surface, like a loose wire or a glitch in the topside control circuits, we try troubleshooting the topside circuit in a rocking boat. No luck. The ROV is still completely unresponsive, and yet both cameras are still rolling and transmitting live video to our TV monitors. Something is definitely wrong, and it doesn’t seem to be the topside. Gulp. Do we do the sensible, conservative thing and abort the dive to recover the cameras (and their SD cards on which the dive’s video is stored) while we still can, even though we’re just shy of our 150 m goal? Or do we risk it all, and hope the cameras will survive long enough to record video at our maximum depth and bring it back intact? Steve gambles that the cameras and camera housings are intact and makes the executive decision to let out more tether and lower the ROV to its full rated depth of 150 meters. We can’t control it, but if the camera housings hold, and if there’s enough light for the cameras to see down there, we’ll at least be able to record something of what's down there. After a few minutes at 150 m, we begin reeling in the 150 meters of tether as quickly as we can. Everyone on the boat is nervous. We get the ROV on board, and the cameras are still running. The camera housing are dry. But there’s no way to turn off the cameras without opening the housings. We don’t want to do that in a rocking boat sloshing with seawater, so we cruise back to our “workshop” as quickly as we can. Long-story short, the cameras and SD cards survived. The electronics can had developed a tiny, slow leak, and some salt water had dripped on the circuit boards, shorting them out and eating way some of the wires.
Day 10: The power has failed on the island. No lights. No ROV battery charging. No air conditioning. We can’t even flush the toilets, because the water pumps don’t work. And, of course, the Ulithi ROV is dead, or nearly so. James is bummed and playing his Ukelele to console himself. The one remaining thing James really wanted to do while we were here was take the (working) ROV down to investigate the wreck of the USS Mississinewa, a US Navy refueling ship that was sunk in the lagoon by a Japanese manned torpedo (called a kaiten) during WWII. Plus, we’re supposed to check some pinnacles near the wreck to get fish population estimates. It’s our last day on Ulithi before we need to start packing everything, and there's no time to repair the ROV, even if we had some power for the soldering iron. Never a group to give up easily, we develop a last-ditch plan. Using the tether and a 100 m transect tape in combination, we decide to turn the Ulithi ROV into a big underwater puppet. We may not have working thrusters or lights, but we do have working cameras, and with these two lines attached, we can raise, lower, and steer the ROV. At least in theory. In the afternoon, we hop in the boat with our disabled ROV and head out to the pinnacles and then to the site of the wreck. The pinnacles turn out to be too shallow to bother with the ROV. Snorkeling suffices to gather the fish observations. Then we move and after a bit of searching, we locate the stern of the wreck upside down in 130 feet of water. It’s a big ship (over 500 feet long), and its upturned belly is barely visible 70 feet below us. It’s impossible to anchor the boat directly over the wreck, so we use an Action Packer tote as a float and a carabiner as a pully to swim the battery box from the boat over to the wreck site and lower it right next to the wreck. While James holds the ROV near the wreck, Steve swims around with the tether tied to the front of the ROV to control the direction it’s pointing. It’s not elegant, particularly since neither James nor Steve can actually see the ROV that far down, but it works. Josh is watching the ROV video live on the TV and is able to guide James and Steve by voice commands. We experience a few delays, like the time the transect tape reel jams and the time the shark comes up out of the blue depths to eye James, but we get some reasonably good video of a sobering piece of WWII history.
Day 11: This morning we started cleaning and drying all the gear in preparation for packing. In the afternoon, we gave a final presentation to the village summarizing our findings during our 12 days on Ulithi. They seemed especially interested in the ROV’s underwater videos, which Josh had edited together into a presentation for them. All day, the villagers were busy preparing a huge farewell feast for us. That night there is much eating, singing, and socializing. Josh, James, and Steve, like many of the men in the village, wear their thus and the traditional headbands made of flowers and ferns. Quite the sight! Josh and Steve spent a good part of the evening chatting with the Chief of Falalop, learning about some of the opportunities and challenges they face, while James plays ukelele and guitar with the local kids. We will miss the wonderful people of Ulithi and this incredible place, but it will also be good to get home and see friends and family again. It’s really weird spending 12 days in a place that has absolutely no contact with the outside world. No TV, no radio, no phones, no newspapers, no internet. We wish we had gotten more ROV data, but we got enough to start a detailed analysis and — hopefully — to prepare for another, more productive visit in a year or two.
Day 12: Departure day. In the morning, we reluctantly make our way back across the island to the airport, which happens to double as the post office. Half the village is there to watch the plane land and take off. It’s the big event of the week. James is playing his Ukelele while some of the Ulithi kids sing along. A tiny white dot appears on the horizon. Soon we can hear the propellers, and a few minutes later, Amos is parking his plane. He walks into the building and switches from pilot to airport ticketing agent. Then he heads back out to the plane as the baggage handler and loads our gear. Next, he’s the flight attendant giving us our safety briefing. A few minutes later, we are airborne and headed for Yap, the first stop on a long journey home.