MASSACHUSETTS INSTITUTE OF TECHNOLOGY Sea Grant College Program

Written by James G. Bellingham for publicaiton in The Global ABYSS: An Assessment of Deep Submergence Science in the United States, University-National Oceanographic Laboratory System, Deep Submergence Science Committee.

Economics provides the fundamental motivation for employing autonomous underwater vehicles (AUV). There are few science missions which can be accomplished better by an AUV than by a manned vehicle or ROV. However, there are a wide range of tasks for which the AUV is either the overwhelming or the only choice for reasons of cost or safety. At present, science AUVs are in the developmental stage. While deep-diving vehicles have been demonstrated, they have not made the transition to operational oceanographic assets. This is not for lack of effort - a number of laboratories both in the United States and abroad are working vigorously on vehicles designed specifically for scientific applications. The most advanced of these new systems are entering field trials in 1994. By the end of the decade, it is likely that AUVs will have become accepted and valued asset to add to the arsenal of deep-submergence science.

Autonomous underwater vehicles, as discussed here, are untethered mobile instrumentation platforms which have actuators, sensors, and on-board intelligence to successfully complete survey and sampling tasks. Unmanned operation without a tether both offers great opportunities and difficult technical challenges. Elimination of the tether both frees the vehicle from the surface vessel and at the same time eliminates the large and costly handling gear which a tether entails. However, because through-water communication is extremely difficult, it imposes the requirement that the vehicle be able to function safely and productively with little or no human supervision.

In principle, the economic benefits of unmanned, untethered operations at abyssal depths have been established through operations of three 6000 meter rated systems: the Advanced Underwater Search System (AUSS) built by the former Naval Ocean Systems Center, the French vehicle, EPAULARD built by IFREMER, and the MT-88 built in the former Soviet Union. However, none of these systems has been adopted by the science community. For the most advanced of these vehicles, AUSS, the reasons underlying this failure to transition stems from high equipment and operational costs, which in turn result from the vehicles technical complexity and size.

In the United States, much of the research and development activity in the AUV area in the past ten years has focused on large systems designed for military purposes. Vehicles weighing in excess of six metric tons have been built. These systems cost millions of dollars to build, and require large support vessels and sophisticated handling capabilities. Also, with the exception of AUSS, abyssal operations has not been a priority, so vehicles have been designed for shallow water operation. While the emphasis on shallow water operations is likely to become even more pronounced with the growing Navy concern with littoral warfare, the dependence on large, expensive systems seems to be changing.

For many years, small vehicles were seen as not having the required payload capacity, range or durability to be useful. However, there have been a dramatic advances in recent years, and a wide variety of institutions are now adopting the philosophy of constructing small, high performance, low cost AUVs. Fundamentally, the objective is to minimize the cost of hardware and operations. The strategy will allow more dangerous missions to be attempted both because the ramifications of vehicle loss are minimized and because under some circumstances the chances for mission success are increased. Two vehicles designed for science missions at abyssal depth are ODYSSEY II, intended for survey operations, and the Autonomous Benthic Explorer which is intended for long duration (up to 1 year) deployment. Both vehicles are undergoing engineering tests in field deployments this year.

Missions and Operational Scenarios

Part of the research and design cycle for AUVs has been the identification of appropriate missions and operational scenarios. There is a temptation to fit new technologies into old roles, however to consider an AUV to be an ROV without a tether, or an oceanographic vessel without a crew, is a mistake. The economic opportunities and technical challenges shaping the emerging class of vehicles are fundamentally different from those of the existing stable of platforms. To take advantage of the unique opportunities posed by such vehicles, new modes of operation have evolved, and starting from the most conservative, are outlined below:

Search and Survey Operations

Of the three most prominent deep-diving AUVs demonstrated to date, AUSS, EPAULARD, and MT-88, all were produced for search and survey operations. Typically such operations would be carried out with deep-towed instrument platforms or sleds. Towed platforms are rugged, are readily modified to carry different payloads, and place few constraints on payload size or power consumption. However, the use of such towed systems is not without its drawbacks. Several factors work in the favor of an AUV:

1. Typical straight-line deep-tow towing speeds, on the order of 1-2 kts, are limited by the drag from the cable as compared to speeds of up to 7 kts demonstrated by AUSS
2. Turning is a time and distance consuming procedure for a deep-tow system, requiring several kilometers for sleds at abyssal depths as compared to turn radii for AUVs that can be less than 10 meters
3. The deep-tow cable and its handling gear imposes a minimum size constraint on a ship as compared to AUVs, of which the smallest abyssal vehicle in development can be operated off of virtually any sea-worthy vessel.
4. The cable and its handling equipment for a deep-tow system can cost as much as the instrument platform
5. Motion of the surface platform is transmitted by the cable to a deep-tow platform, which is a particular problem for applications in which stealth or accurate positioning is critical (e.g. mid-water column biology)
6. Accurate positioning of deep-towed systems requires a surface vessel with dynamic positioning

The Naval Ocean Systems Center in San Diego estimated a three to four-fold increase in search rates of their AUV as compared to deep towed systems. Soviet operations with the MT-88 demonstrated that even with equal speeds, a 4 kilometer square area at 5000 meters can be surveyed two to three time more rapidly with an AUV than with a towed system due to the more maneuverable characteristics of an AUV. Thus provided issues of handling, reliability, and equipment costs can be adequately addressed, deep search and survey operations can be significantly enhanced by use of AUVs.

Complementary Use

One promising operational scenario is to use one or more AUVs to complement ROV and manned vehicles operations. A wide variety of activities require either the physical presence or telepresence of a human, especially tasks which are vision intensive and/or require manipulation. However, both ROVs and manned vehicles have limited mobility compared to many AUVs. This is suggests the following uses of AUVs to increase the effectiveness and economy of ROV and manned vehicle operations.

Manned submersible operations provide ocean researchers with unparalleled opportunities to observe the marine and benthic life forms, environments, and geology. However, the amount of actual bottom time and distance traveled per dive is usually quite limited, so routine geophysical or water column surveying by these submersibles, although possible, is generally not an efficient or economical use of their capabilities. AUVs offer the capability to obtain good lateral coverage. A relatively low cost AUV could be used to locate scientifically interesting sites for later visitation by an ROV or manned vehicle. Thus an Alvin dive, for which the average bottom time is three to four hours, could be utilized to much greater effect in a previously unexplored area if preceded by an AUV survey. Since Alvin dives occupy only about half the ship time of the support vessel, the unused ship time is available for AUV operations with no impact on ALVIN operations. Such a scenario has been suggested by D. Fornari, and will be pursued on an ALVIN cruise late in 1994.

Simultaneous operations of AUVs and ROVs are also attractive. For example, ROVs are not ideal for large scale lateral surveys in deep water because the tether dominates the dynamics of a typical vehicle-tether system for tether lengths longer than about 1000 meters. Physically moving an ROV deployed to abyssal depths means that many close observation and sampling activities are precluded. A small survey class AUV, in contrast, could operate independent of a support vessel for short periods and thus allow lateral exploratory probes. The ROV could thus remain on a site of interest, and be moved only on discovery of a more scientifically interesting location by the AUV.

A second use of AUVs to complement ROVs is the "elevator" application. ROVs are capable of operating continuously on the bottom for many days, but for many applications are forced to the surface at frequent intervals to return samples. Biological and chemical operations in particular often require timely return of specimens and samples. However, recovering an ROV to the surface and subsequently redeploying it to depth can consume a substantial fraction of a day, significantly degrading the efficiency of the system and excessively exposing it to the risks incumbent in deployment and recovery. The elevator AUV would eliminate the need to recover the ROV by providing a shuttle system for returning samples to the surface.

Rapid Response

Rapid response missions for AUVs capitalize on their potential for both low cost and ability to operate from ships of opportunity. Fundamentally, the scenarios hinge on the assumption that a deep-ocean platform must be held in reserve to respond quickly to some unpredictable event. The economic equation revolves not only around the platform, but also around the supporting oceanographic vessel. Manned submersibles typically operate off dedicated support vessels. ROVs and deep-towed systems, while less demanding in their support vessel requirements, still need a large oceanographic vessel with appropriate winches and dynamic positioning capability. Consequently, not only must the instrument platform be held in reserve, but an oceanographic vessel as well. In contrast, the smaller AUVs can be operated off of virtually any ship of opportunity, or possibly even deployed from an aircraft.

Volcanic eruptions on spreading ridges provide an example of an episodic event of interest to science. Such events are important in terms of the geophysics and geology of the seafloor, the biological communities of the seafloor, hydrothermal vents, and the lower water column, and the heat and chemical balance of the ocean. Considerable interest has been expressed in the scientific community (see RIDGE meeting reports) to study the aftermath of such an eruption, and, if at all possible, to study an event while in progress.

As a step towards enabling a rapid response to episodic events on spreading ridges, NOAA PMEL has demonstrated remote detection using the Navy SOSUS acoustic array. To capitalize on this, MIT Sea Grant and NOAA PMEL are working to demonstrate that the ODYSSEY II AUV can be used for surveying operations to demonstrate a rapid response capability. This combination of technologies should allow scientists to detect and respond in an economic manner to eruptions in progress at active spreading centers.

Observatories

The abyssal observatory provides an emerging class of facilities for making measurements in the deep ocean. The motivation for such observatories includes: a desire to coordinate science programs in different disciplines to promote an interdisciplinary approach to ocean processes, the need to coordinate time-series studies to obtain spatial coverage, and the opportunity to obtain real-time data from instruments on the sea floor. Many of these motivations were discussed in the previous chapter in the context of the Ridge Observatory Experiment (ROBE). While existing observatories have not incorporated AUVs, efforts to do so are in progress.

Observatories take many forms. One example is the Acoustic Local Area Network (ALAN) set up in Monterey Bay, which consists of a surface buoy which has a radio link to shore and an acoustic modem link with an array of scientific instrument packages distributed in the vicinity of the surface buoy. In contrast, the LEO-15 mooring is going to have a cable link with shore, which will provide both communication via a fiber-optic link and power transmission. While LEO-15 is a shallow water system, similar deep-water moorings have been proposed. A characteristic in common to these approaches is the communication link with shore.

The Autonomous Benthic Explorer (ABE) is an AUV designed to address the need for obtaining spatial coverage of a defined area, for example the immediate vicinity of a hydrothermal vent, over a period of many months. Described more fully below, ABE is intended to operate from a near-bottom mooring from which it will detach itself periodically and fly preprogrammed surveys. At the end of a survey, ABE will reattach to the mooring and shut down for power conservation until the next survey. The data so obtained is stored on ABE until the vehicle is recovered at the end of its deployment period. The SEAFLOOR ROVER, a crawling vehicle described later in this chapter, has a similar long-deployment nature, although because it is not free-swimming, it does not need to secure itself to a mooring between periods of activity.

An approach which incorporates elements from all of the systems described above is the Autonomous Ocean Sampling Network (AOSN) concept, in which moored buoys provide power and communication nodes to provide a long term, multiple vehicle presence in the ocean. The objective is to provide an economically feasible capability for repeated synoptic characterization of large scale oceanographic phenomena. The key to such a system is a small, low-cost autonomous vehicle which can be operated reliably over extended unattended deployments at sea. The motivation for multiple vehicle surveys is that the quality and utility of the data obtained improves much faster than the number of vehicles for large scale, dynamic ocean phenomena. To the scientist using the network to study the ocean, the instruments appear as Internet site(s) - the many different physical links (satellite, acoustic modem, etc.) are transparent to the user. With an AOSN it should possible to monitor real-time data from the field, and to redirect the AUVs collecting data to optimize their productivity, thus greatly extending our presence in the ocean in an economical fashion.

<- top of page ->