Home | Contact ST  

Feature Articles

Waitt Institute Completes Pacific Survey Utilizing AUV Technology
REMUS AUVs Explore Deep Reefs, Search 2,200 Square Miles Of Seafloor for Amelia Earhart’s Missing Aircraft


Michael Dessner
Director of Operations
Waitt Institute for Discovery
La Jolla, California

Greg J. Packard
Senior Engineering Assistant
Applied Ocean Physics and Engineering
Woods Hole Oceanographic Institution
Woods Hole, Massachusetts

Andy Sherrell
Senior Ocean Engineer
Harbor Branch Oceanographic Institute
Florida Atlantic University
Fort Pierce, Florida
In the fall of 2008, the Waitt Institute for Discovery took possession of two Hydroid Inc. (Pocasset, Massachusetts) REMUS 6000 autonomous underwater vehicles (AUVs) and began a collaboration with Woods Hole Oceanographic Institution (WHOI) and Harbor Branch Oceanographic Institute (HBOI) at Florida Atlantic University.

The collaboration would test the equipment at full operational depth, undertake initial research to ground truth deep reefs off the coast of Florida and prepare for a survey off Howland Island in the central Pacific. The Pacific mission was designed to locate the Lockheed Model 10 Electra Amelia Earhart was flying in 1937 when she and Fred Noonan disappeared en route to Howland Island.

Although the collaboration did not find Earhart’s plane, its overall efforts were successful: Sea trials resulted in reliable operations, the initial science work found and mapped several deep Lophelia and Oculina reefs, and the Earhart search produced a 2,200-square-nautical-mile mosaic of high-resolution side scan data.

A three-meter resolution multibeam image of three deep reefs, named Triceratops, discovered by the initial expedition working with Harbor Branch scientist John Reed.
Teams and Equipment
The two Hydroid AUVs are completely autonomous vehicles with ground truthing capability and are primarily used as a bed for deep sensing payloads. The AUVs are equipped with an inertial navigation system capable of working in conjunction with long baseline acoustic transponders, allowing for an exceptionally high degree of navigational accuracy as the vehicle travels in and out of the long baseline’s acoustic footprint. The primary sensor is a 120/410-kilohertz dual-frequency side scan sonar. The system may also be fitted simultaneously with a sub-bottom profiler, multibeam sonar or an electronic still camera, which is used for visual confirmation of contacts. The vehicles utilize a forward-looking pencil beam sonar along with software algorithms to provide an effective and autonomous collision avoidance system.

The Waitt Institute has been focused on the search for Amelia Earhart for more than five years, culminating with two deep-sea expeditions from November 2008 through May 2009. The institute entered into collaboration with WHOI to create a working group to operate the AUVs during Waitt Institute missions. The Waitt Institute also partnered with HBOI during this period for project management, additional operational and scientific personnel for the two expeditions, and the use of the platform from which the expeditions took place, the 1,300-ton, 202-foot RV Seward Johnson.

Trials and Initial Science
The initial shakedown of the AUV systems took place in November 2008, 220 miles off the east coast of Florida in 1,200 meters of water. Collaborators tested the long baseline navigation system, conducted launch and recovery procedures for the deep-ocean trans-ponders and operated both vehicles at depth while developing procedures for extended, efficient and concurrent vehicle operations. Operating two vehicles simultaneously in deep water is something that had not been previously attempted by private users of REMUS AUVs. Simultaneous operation proved challenging, but the outcome increased the overall efficiency of the expedition almost twofold, as procedures were developed to seamlessly transition between vehicles without negatively affecting their preplanned missions.

The expedition then returned to Harbor Branch to change out crew for the second leg of the shakedown phase, which included deeper operations down to 5,500 meters and allowed Harbor Branch senior scientist John Reed to attempt to find and ground truth deep reefs. The AUV side scan sonars successfully detected and imaged several deepwater Lophelia and Oculina reefs. The vehicles were then equipped with high-resolution multibeam sonars to map these newly discovered reefs in full relief. The results were submitted to federal fisheries regulators, who used the data to recommend creating the largest marine protected area along the Atlantic coast. The expedition was covered in depth by public television and reported on the Harbor Branch and Waitt Institute websites.

A final segment of sea trials was added to test the vehicles at full operational depth. Due to weather constraints, this work was conducted in five days, 550 miles offshore in 3,000 meters of water to ensure that near full-depth, full-length concurrent operations were feasible. Operational refinements led to the development of procedures that would be used over the course of the upcoming Pacific mission. After three days of concurrent vehicle operations at depth, the ship returned to its home berth in Fort Pierce, Florida, for full provisioning. The teams returned home for the holidays and the timeline was finalized for the search for Earhart’s airplane.
The Search for Amelia
In February 2009, teams from WHOI, HBOI and the Waitt Institute flew to American Samoa to board the Seward Johnson for transit to the search area, located 1,110 miles north-northeast of the expedition’s operational base in Pago Pago, the capital of American Samoa.

The search area was drawn from a report produced by experienced aircraft accident investigators who had spent months analyzing Earhart’s 1937 disappearance. After poring over thousands of documents, performing cutting-edge simulations and inspecting extant Lockheed Electra 10 series aircraft, investigator Chris Nutter and his associates arrived at three potential end-of-navigation points. The results of that report were used by the investigators and the authors of this article, in collaboration with Ted Waitt, the founder of the Waitt Institute for Discovery, to create a 2,500-square-nautical-mile search area.

The initial search grid, subsequently modified due to theoretical and operational constraints, was divided into four columns, each of which contained 22 discrete boxes representing a single vehicle mission. This grid stood on a vertical axis running south-southeast to north-northwest on the western side of Howland Island. The expedition had two 45-day missions to map the area using high-resolution side scan sonar.

Initially, the AUVs were outfitted with EdgeTech (West Wareham, Massachusetts) dual-frequency 75/410-kilohertz side scan sonars. Over the course of the expedition, they were replaced with more robust EdgeTech 120/410-kilohertz transducers. The vehicles began survey work on the easternmost two columns by deploying deep-ocean transponders at the vertices between the boxes and progressing up the “ladder” of the search, advancing northward.

Both vehicles were operated simultaneously. Deep-ocean transponder launch and recovery points were developed ad hoc to allow for mutability of operations should the need arise due to malfunction or the desire to remain in the area to investigate targets.

In each discrete search box, an AUV was programmed to run a staggered line pattern at a 600-meter-range scale with 450-meter and 1,150-meter line spacing. This provided 100 percent coverage of the gap with a 50-meter overlap. HYPACK Inc. (Middletown, Rhode Island) software was used for planning and mapping purposes and 20-20 software by Highland Solutions Inc. (Tualatin, Oregon) was used for side scan interpretation and analysis. The vehicles operated at an average depth of 5,200 meters and an average speed of 3.5 knots.

Each discrete search box was roughly 24 square miles. From deployment to recovery, it took a single vehicle approximately 23 hours to cover a search box, during which it would average 18 hours of bottom time. Upon recovery, crews changed the batteries and prepped the vehicle for its next mission while sonar analysts quickly reviewed the acquired data to determine the presence of targets of opportunity that the vehicles might be programmed to reacquire on the following sortie. Minimizing time out of the water for the AUVs was critical to maintaining the highly efficient operational pace that the expedition set as a benchmark; vehicle turnover time from recovery to launch averaged less than four hours.

Target reacquisition was accomplished by programming the vehicle to interrupt its next mission as it was working the subsequent box. When the vehicle neared the area of the intended target, it would follow preprogrammed instructions to break off from the survey, navigate to the target area, drop in altitude, switch side scan frequencies to the higher resolution channel and execute a tight grid search over the target while also firing its electronic still camera. Generally this would provide the analysts with multiple short-range, high-resolution sonar images of the target, along with some 3,000 to 5,000 photos of the area.

Once the process was complete, the vehicle would fly back to the original survey and reset its altitude and sonar frequency to complete the original mission. Typically this maneuver would only take a few hours to complete and, upon recovery, provided a clear and deterministic view of the object of inquiry. This proved to be an invaluable operational capability. Other systems used at these depths, typically towed arrays, require up to 12 to 24 hours to reimage targets in higher resolution sonar and, in many cases, would not conclusively identify the target. The REMUS AUVs’ ability to ground truth eliminates the need for costly and time-consuming follow-up expeditions to identify targets using remotely operated or human occupied vehicles.

Each vehicle performed more than 1,000 hours of service during the expedition, traveling a combined total of more than 7,000 linear miles over the seafloor and generating 2,200 square miles of high-resolution side scan sonar mosaic. This operations phase was completed in 72 days on site and includes the time it took to reimage all targets by scanning or photography.

Aside from geologic formations, the expedition detected two coring pipes (or something similar) standing like flagpoles on the bottom, a pipe or chain lying on the bottom, a 55-gallon drum and a transoceanic cable that tracked more than 30 miles across the survey area.

Target reacquisition successes combined with the ability to photographically ground truth sonar contacts resulted in high confidence that the failure to find the plane was not a failure to detect. The authors are confident that this data, added to the survey conducted by Nauticos and the Waitt Institute for Discovery in 2006, creates a 2,700-square-mile zone of exclusion inside the long-held 5,000-square-mile area of Earhart’s potential disappearance.

No equipment losses or personnel injuries were sustained during the expedition.

Analysts processed data during transit and subsequent to the mission, creating an exceptional mosaic of the area surveyed.

The Waitt Institute, in conjunction with its partners on this search, has made all data accessible to the public through the Marine Geoscience Data System and GeoMapApp, a data exploration and visualization tool developed by the Lamont Doherty Earth Observatory. Data from all shipboard automatic monitoring systems are also available to the public through the same tool, as well as the results from biological survey net trawls during the expedition and data from the vehicle’s conductivity, temperature and depth sensor.

All research that led to the search, additional photographic images of targets and other multimedia are available at the Waitt Institute for Discovery’s Search for Amelia website.

Michael Dessner, director of operations for the Waitt Institute for Discovery, has worked as a manager in the ocean industry for the last 20 years. His background includes extensive experience in Alaskan commercial fisheries, among other ocean-related enterprises. He was invited to join the institute by Ted Waitt in 2005.

Gregory Packard, senior engineering technician within the Oceanographic Systems Lab at Woods Hole Oceanographic Institution, holds a B.S. in marine transportation from the Massachusetts Maritime Academy. Packard was instrumental in the engineering and development of the REMUS AUVs, heading up numerous underwater vehicle field exercises and training personnel from the military, science and private sectors in vehicle operations and applications.

Andy Sherrell holds a bachelor’s degree in ocean engineering from the Florida Institute of Technology and currently works as a senior ocean engineer at Harbor Branch Oceanographic Institute at Florida Atlantic University in Fort Pierce, Florida. His 15 years of ocean experience includes numerous deep-sea search and salvage operations with in-depth expertise locating and recovering downed aircraft.

-back to top-

-back to to Features Index-

Sea Technology is read worldwide in more than 110 countries by management, engineers, scientists and technical personnel working in industry, government and educational research institutions. Readers are involved with oceanographic research, fisheries management, offshore oil and gas exploration and production, undersea defense including antisubmarine warfare, ocean mining and commercial diving.