R. Camilli, MIT Parsons Laboratory Building 48-202, 77 Massachusetts Avenue Cambridge, Massachusetts 02139

Kemonaut is an Autonomous Underwater Vehicle based on the MIT Sea Grant Odyssey II design, incorporating a free flooding hull with three internal glass pressure sphere housings, allowing the NEREUS submersible membrane inlet mass spectrometer to be carried as payload in its forward bay. The vehicle-payload combination weighs approximately 200 kg in air and is designed for operation in freshwater and marine coastal environments to a depth of 300 meters. Composite structural components, including carbon fiber laminates and high-density structural foam, are used throughout the Kemonaut vehicle and NEREUS instrument to minimize weight while maintaining adequate strength and vehicle stability. Stability is further increased through asymmetric positioning of the glass pressure spheres relative to the vehicle centerline. This design permits greatly expanded vehicle payload capacity while preserving the Odyssey II hull shape and allowing for the use of existing propulsion and control systems.

Sea trials have successfully demonstrated the vehicle and instrument configuration's unique ability to autonomously measure multiple dissolved atmospheric gases at ambient concentrations throughout the water column. Applications particularly well suited for the Kemonaut-NEREUS configuration include pollution identification and mapping to aid in remediation efforts, assessment of sub-surface natural resources, and research on marine-related greenhouse gas cycling.

MIT Sea Grant's Odyssey Class autonomous underwater vehicles (AUVs) are small, easily deployable, low component cost survey platforms, which have been used in numerous missions throughout the world (Bellingham 1997). These vehicles make use of low-cost mass produced hollow glass spheres, which commonly function as full ocean depth floats, to house vehicle electronics, sensors and power systems in a dry one-atmosphere environment, thereby permitting the use of a free flooding vehicle hull and minimizing vehicle volume and displacement requirements. However, the minimization of vehicle displacement makes the tradeoff between range (i.e. power storage) and payload capacity a critical design consideration (Bellingham, Goudey et al. 1994). The NEREUS membrane inlet mass spectrometer, which is highly optimized to handle the difficult hotel function constraints imposed by small autonomous underwater vehicles, is completely self contained within an Odyssey compatible 17” glass pressure sphere, weighs 22 kilograms, and operates to depths of 100 meters on a self contained power supply, drawing less than 20 watts (Hemond and Camilli 2002). NEREUS uses an embedded autonomous control system capable of adapting its mission directives and sampling regimes to better monitor its environment. Operation of the NEREUS instrument onboard an autonomous underwater vehicle enhances the advantages of in-situ analysis of dissolved volatile chemicals by increasing the range and overall accessibility of data collection areas. Furthermore, unlike other platforms such as moorings, tow fish, and remotely operated vehicles, an AUV does not require a tether, thus allowing a spatial survey to be conducted faster, follow bathymetry better and operate in high sea states.

In August 2002, after having undergone successful limnologic and marine trials as both a buoyed and towed sensor, the NEREUS instrument was ready for integration into an Odyssey class AUV. At this time MIT Sea Grant's single remaining Odyssey II Xanthos vehicle (Figure 1) used both of its internal pressure spheres to house vehicle power, control, and communication systems, making it physically impossible to accommodate the NEREUS instrument without consolidation of the components into a single sphere. NEREUS deployment aboard MIT Sea Grant's newer

Sea Grant's Odyssey Class autonomous underwater vehicles (AUVs) are small, easily deployable, low component cost survey platforms, which have been used in numerous missions throughout the world (Bellingham 1997). These vehicles make use of low-cost mass produced hollow glass spheres, which commonly function as full ocean depth floats, to house vehicle electronics, sensors and power systems in a dry one-atmosphere environment, thereby permitting the use of a free flooding vehicle hull and minimizing vehicle volume and displacement requirements. However, the minimization of vehicle displacement makes the tradeoff between range (i.e. power storage) and payload capacity a critical design consideration (Bellingham, Goudey et al. 1994). The NEREUS membrane inlet mass spectrometer, which is highly optimized to handle the difficult hotel function constraints imposed by small autonomous underwater vehicles, is completely self contained within an Odyssey compatible 17” glass pressure sphere, weighs 22 kilograms, and operates to depths of 100 meters on a self contained power supply, drawing less than 20 watts (Hemond and Camilli 2002). NEREUS uses an embedded autonomous control system capable of adapting its mission directives and sampling regimes to better monitor its environment. Operation of the NEREUS instrument onboard an autonomous underwater vehicle enhances the advantages of in-situ analysis of dissolved volatile chemicals by increasing the range and overall accessibility of data collection areas. Furthermore, unlike other platforms such as moorings, tow fish, and remotely operated vehicles, an AUV does not require a tether, thus allowing a spatial survey to be conducted faster, follow bathymetry better and operate in high sea states.

In August 2002, after having undergone successful limnologic and marine trials as both a buoyed and towed sensor, the NEREUS instrument was ready for integration into an Odyssey class AUV. At this time MIT Sea Grant's single remaining Odyssey II Xanthos vehicle (Figure 1) used both of its internal pressure spheres to house vehicle power, control, and communication systems, making it physically impossible to accommodate the NEREUS instrument without consolidation of the components into a single sphere. NEREUS deployment aboard MIT Sea Grant's newer Odyssey III Caribou vehicle was unlikely until spring of 2003 because vehicle dynamics using a NEREUS payload section in combination with the vehicle's vectored thruster and MOOS operating system were not well understood and because delivery of a payload section from Bluefin Robotics was estimated to require several months.

Figure 1: Digaram of Odyssey IId Xanthos AUV (courtesy of MIT Sea Grant)

After lengthy analysis it was decided that a new type of Odyssey vehicle could be built specifically for NEREUS missions at an economic and time cost lower than that of the Caribou center section or the Xanthos reconfiguration. Design of this new vehicle, named Kemonaut, began in late October 2002 with the goal of producing an operational vehicle before winter weather delayed NEREUS deployment until the following spring. To speed the development process, it was determined that the Kemonaut design should use an external hull, tail cone, control and power system spheres identical to the Odyssey IId (Manley and Rieffel 2000). However, because the Kemonaut vehicle required space for three glass pressure spheres (two for standard Odyssey IId power and control and one for the NEREUS instrument) while retaining the same external dimensions of an Odyssey II hull, a number of new design factors had to be addressed, including:

• increased vehicle displacement
• decreased hull stiffness
• decreased volume for core vehicle components
• decreased volume for floatation foam
• unknown vehicle center of buoyancy, center of mass, and dynamics
• unknown navigation and communication system interference from the NEREUS instrument

To further speed the design process, Kemonaut design plans were developed on a computer using Solid Works CAD software, allowing analysis work to be undertaken before vehicle components were built and thereby minimizing the excessive time and economic costs of iterative physical fabrication and construction cycles.

The design process began by first obtaining an Odyssey II outer hull fairing From MIT Sea Grant's AUV Laboratory. The fairing's diameter was then measured at 30 centimeter transect intervals along its longitudinal axis. These dimensions were then used to loft a computer model of the external hull shape, which served as relative boundary volume for vehicle components. Three 17” diameter spheres were then described within this internal volume, yielding a residual volume for all other vehicle components. To further increase clearance between the 3 spheres, as well as overall vehicle stability (lowering the vehicle center of mass) the glass pressure spheres were asymmetrically positioned relative to the vehicle centerline (Figure 2). In this arrangement the heaviest sphere containing the vehicle battery supply was located in the lower central position, the NEREUS sphere was located in the forward position, and the vehicle control sphere was positioned in the aft. This arrangement provided several benefits, including:

• Additional volume in the upper section of the hull is made available for floatation foam, further improving stability characteristics.
• Spacing between the vehicle's navigation and communication systems and the NEREUS instrument is increased, thereby minimizing electro-magnetic interference between vehicle and NEREUS systems.
• Distance between the two vehicle spheres and the tail cone is minimized, thereby minimizing required cable length.
• NEREUS sampling occurs before water is perturbed by the vehicle's motion.

Despite the advantages of this arrangement, the layout revealed that large sections of the Odyssey II structural inner fairing would need to be removed in order to accommodate the third NEREUS sphere, dramatically compromising the strength of the inner hull. Furthermore, the additional 48 kg NEREUS sphere would require greater hull strength than a standard twin sphere Odyssey II layout.

Figure 2: Glass pressure sphere placement within the Kemonaut AUV

To prevent hogging or more severe structural failure leading to possible pressure sphere damage during launch and recovery, an additional hull girder was essential for improved hull rigidity. However, the third sphere and an additional girder significantly reduce available volume for floatation foam, meanwhile requiring supplementary buoyancy to offset the increased displacement produced by these new components.

Similar strength and mass constraints had previously been confronted during development of the NEREUS instrument, leading to the use of composite laminates and foams in its structural elements. These materials functioned well in that they: reduced component weight by approximately 80%, endured high shock loads without failure during instrument deployments. In addition, these materials are electrically and magnetically non-conductive (a consideration when using fluxgate sensors), are generally immune to corrosive sea water, and can be fabricated into complex shapes relatively quickly and at low cost. Based on the positive results met by the composite NEREUS components, a computer design of a composite longitudinal girder was developed to fulfill the structural requirements of the Kemonaut vehicle. The contours of this surfboard-shaped component allow it to occupy the normally empty four-inch gap, running the length of the hull, between the upper and lower Odyssey II inner fairings. To further reduce the vehicle's floatation foam requirement, this girder design utilizes marine structural foam (? = 288 kg/m3) sheathed in a carbon fiber cloth (? = 1,570 kg/m3). The structural foam itself possesses flexural strength (rise parallel to beam thickness) of 4.8 MPa, while the autoclave cured carbon material possesses a tensile strength of 448 MPa. Because the NEREUS instrument is intended to operate to depths of approximately 100 m, the structural foam was chosen as a compromise between the high buoyancy but depth limited low-density foams and the enormous depth capability of less buoyant syntactic foam. The carbon fiber is an autoclave cured bi-directional (0°/90°) 33 msi 2x2 twill, layered to 1.5 mm thickness, which is impregnated with an epoxy binding matrix, providing a measure of armor plating for the foam as well as a large proportion of the girder's rigidity.

Construction entailed creation of a plywood template to pattern the contours onto the carbon cloth and structural foam. These materials were then cut to shape using a high-speed spiral saw. After trimming, the upper and lower carbon fiber sections were bonded to the central foam section and vacuum cured using a small rotary pump and simple vacuum bag constructed of trash bin liners. Autoclave curing of the carbon/foam structure was not attempted, as it would have been deleterious to the strength of the foam. External sides of the foam piece were also sheathed in carbon cloth, but were bonded using an epoxy containing glass microfibers to decrease brittleness and thereby minimize the likelihood of delamination. Curing was accomplished using the previously mentioned vacuum bag technique. Cutouts for vehicle components were then made to the girder after curing was completed. The resulting composite structure has a mass of approximately 9 kg and is calculated to withstand depths to 300 meters without significant compression taking place and, due to the density of the foam, provides approximately 16kg of floatation.

Figure 3: Exploded diagram of Kemonaut vehicle components

After fabricating the girder, the lower inner hull fairing was modified to accept the third NEREUS sphere by excising two sphere cradles from a surplus Odyssey II inner hull and repositioning these further apart within a second Oddysey II inner hull that was trimmed to accept the transferred cradles. The third, central, sphere cradle was constructed of a non-ribbed polyethylene Benthos “hardhat”. Accurate positioning of the sphere cradles was necessary because of the limited clearance between the spheres. Therefore, to insure proper spacing, glass hemispheres were placed in the cradles and aligned using the composite girder before the sphere cradles were riveted into place with stainless steel pop rivets.

A double-point lifting harness was designed instead of the single-point system that Odyssey II vehicles use for multiple reasons. First, adequate clearance was not available for a single central lift point because of the location of the center sphere. Second, two relatively wide-spaced lift points would cause the vehicle's center of mass to reside between the two points, allowing the vehicle to be lifted without the vehicle experiencing the uncontrolled seesawing common to the single-point harness. Third, by distributing the loading between two attachment points, failure was less likely to occur. After finalizing a computer design, the lift frame was fabricated of 1/8”x 1” cross section annealed #316 stainless steel bar stock and 1” diameter stainless steel eyebolts.

Placements of those components residing outside of the spheres were decided based on orientation requirements for proper operation and computer generated clearance estimates. As such, the sonar altimeter was positioned vertically between the forward and middle sphere on the vehicle's starboard side and a cutaway was made in the underside of the lower hull to expose the altimeter's acoustic head. To help equalize weight distribution, placement of the depth sensor mirrored the sonar altimeter along the central axis, residing on the portside between the forward and middle sphere. Careful consideration led to vertical placement of the GPS Intelligent Buoy (GIB) transponder in the nose of the vehicle, permitting the transponder head to sit proud of the hull for clearer signal transmission. The battery housing supplying power to the strobe beacon was placed immediately forward and starboard of the tail cone assembly in another attempt to aid balance by offsetting the asymmetric girder cutaway required for the tail cone assembly. The strobe light beacon, GPS antenna, and radio antenna were mounted immediately aft of the lift harness' rear eyebolt on top of a 3.5 inch diameter cylindrical long-base-line (LBL) transducer that had been permanently affixed to the upper hemisphere of the Xanthos control sphere.

A modular three-piece floatation system was designed to mount within the upper half of the vehicle between the middle sphere and outer hull to increase vehicle payload carrying capacity and further improve stability. Six pieces of 3 1⁄2” thickness pressure resistant marine foam (identical to material used in the composite girder) were cut to track the hull shape using dimensions derived from clearance estimates between the outer hull fairing, lift harness, and pressure spheres. The four forward-most pieces were then bonded together and coated with an epoxy to improve durability, whereas the remaining two foam pieces were coated with epoxy and cured separately. The large 4-piece section provides approximately 15 kg of buoyancy with an additional 2 kg of buoyancy available through the inclusion of each of the small single sections, thereby permitting buoyancy adjustment according to vehicle-payload configuration. This modular system, when combined with the composite girder, provides between 31 and 35 kg of gross vehicle buoyancy when deployed in freshwater.

Vehicle control was made possible using the Odyssey IId Xanthos tail cone assembly and control sphere. Both components are powered using the central battery sphere which houses a battery pack capable of generating 800 watt-hours of energy. The tail cone assembly includes a propulsion unit powered by a 100 watt Pittman ELCOM series 5100 brushless thruster enclosed within a cylindrical metal Benthos pressure housing in concert with an elevator and rudder subassembly composed of two oil compensated Pittman 3100 actuators individually directed by JR Kerr motor controllers using Hall effect feedback sensors (Damus, Desset et al. 2002).

Navigation is carried out through a simple open-loop system, using a global positioning system and CrossBow AHRS (altitude, heading and reference system composed of a 3-axis accelerometer and magnetometer) contained within the control sphere, along with a Data Sonics PSA-916 sonar altimeter and Paroscientific Digiquartz pressure sensor (Damus, Desset et al. 2002). Data from these sensors are supplied to the main vehicle computer (MVC), a 300 MHz PC-104 stack, which uses MIT Sea Grant's Mission Oriented Operating System (MOOS) and is located within the aft control sphere. A FreeWave spread spectrum radio modem housed in the aft sphere supplies communication between the MVC and base station. This basic configuration allows the vehicle to obtain position fixes and mission updates by way of the radio transmitter and onboard GPS when the vehicle is surfaced. While submerged, the MVC executes mission directives using dead reckoning estimates via the AHRS, altimeter and pressure transducer. The MVC automatically logs time stamped MOOS data during both surface and submerged operation, permitting post-mission analysis of vehicle operations and time synchronized integration of vehicle track log data with NEREUS chemical data.

After fabrication was complete, the new 3-sphere inner structural hull was fastened to the outer hull using stainless steel screws. The tail cone assembly was then fitted to the external hull, using a laser to align the thruster axis with the vehicle centerline, and held in place with radially bored stainless steel screws. Next the sonar altimeter and pressure transducer were mounted in the lower hull with hose clamps banding their housings to stainless steel L-brackets that are in turn bolted to the inner hull. Desiccant packs were added to the pressure spheres before sealing to prevent temperature-induced condensation on internal electronic surfaces, then evacuated to approximately 0.8 atmospheres and placed within their appropriate inner hull cradles. A small hull cutaway was made in starboard bow to accommodate the NEREUS instrument's sampling inlet and a low-cost engine coolant water intake guard was then bolted in place over the cutaway to act as a guard for the sampling inlet. Electrical wiring residing outside of the pressure spheres were connected using waterproof wet pluggable MIL-9 and coaxial cables. The composite girder was then positioned on top of the lower inner hull, followed by the upper inner hull, floatation foam and lift harness, in respective order. These components were then secured in place by eight stainless steel bolts, placed vertically along the perimeter, with the four central bolts attaching to the lift harness legs for load transfer, buttressed by two pairs of bolts fore and aft to prevent hull separation. The top outer hull was positioned and secured along its edge using Zeus fasteners once all components were in place. Finally, several small holes were drilled in the hull underbelly to permit adequate water drainage during vehicle recovery.

Figure 4: Photograph of partially assembled Kemonaut,
showing the NEREUS instrument and composite girder

Prediction of Kemonaut's displacement and ballasting requirements were made prior to construction using computer-generated models of the assembled vehicle. The overall vehicle was calculated to have a mass of nearly 185 kg and displace some 0.196 m3 volume, thus yielding a net buoyancy of about 15 kg in freshwater by assigning appropriate material density characteristics to individual Kemonaut components. Modeling and calculations also indicated that proper trim could be achieved by placing 10 kg of lead weight in the aft portion of inner hull immediately in front of the tail cone assembly, along with approximately 4 kg of lead below the forward sphere, producing a net 1 kg positive buoyancy in freshwater with a center of mass located approximately 6 cm below the center of buoyancy.

Initial ballasting was carried out on December 10, 2002 in the Upper Mystic Lake after the vehicle had been assembled. At this time the lake surface had frozen to a thickness of several centimeters, requiring that a hole be made in the ice for ballasting. Using the Mystic Boat Club's facilities, the vehicle was lowered by electric hoist into the water to observe static buoyancy. Lead weight was then added iteratively until the vehicle established approximately 1 kg positive buoyancy and relatively neutral trim. A total of 15.4 kg of lead weight was added to the vehicle, with 8.8 kg (two 10 lb lead dive weights) placed immediately forward of the tail cone, 4.4 kg (one 10 lb dive weight) between the forward and middle inner hull cradles, and two 0.9kg (2 lb dive weights) on the port and starboard beam each. Based on this ballasting, the vehicle was calculated to need between 3.6 and 5.4 kg of additional ballast weigh for marine deployment.

Figure 5: Kemonaut ballasting in Upper Mystic Lake

On December 19 2002 the Kemonaut vehicle and NEREUS payload underwent sea trials in Boston Harbor to demonstrate the cargo carrying capacity, maneuverability, and overall seaworthiness of the Kemonaut design as well as the ability of the NEREUS instrument to collect meaningful chemical data while operating onboard a moving AUV platform. A relatively confined region of Boston's Inner Harbor near to the Charles River Locks was selected for operations because of its unobstructed flat bathymetry, relatively low frequency of boat traffic during the winter months, and absence of underground tunnels or electrical cables thereby avoiding potential electromagnetic interference to the vehicle navigation system.

Operations began with the deployment of four GIB buoys with attached hydrophones around the perimeter of the test area. These buoys were positioned in more or less a rectangular pattern to maximize the triangulation distance while minimizing propagation path obstructions and interference with shipping traffic. Rodes were anchored in place using a 5 kg weight attached to 2 meters of 3/8" steel chain to prevent drift in the channel's +1 kt current and 3.5 meter tidal cycle.

Boston Harbor's Pier 4 served as the operations base station for radio communication with Kemonaut and crew aboard a skiff that served as an intercept boat. Operations base station was setup in the cargo bed of a Ryder truck, parked at the end of the pier to provide a clear view westward out over the Harbor.

Kemonaut was ballast tested for marine deployment from a two-ton electric hoist once GIB transponders were in position and functioning. Based on results from earlier freshwater ballasting, an additional 3.8 kg of weight had been fitted to the vehicle while undergoing assembly in the lab. This additional weight was distributed in two equally weighted 1.8 kg bags of lead shot. The first bag was zip-tie wrapped to the 10 lb dive weight in the nose of the vehicle, and the second weight was wired tightly to the external aft surface of the vehicle, directly below the two 10 lb dive weights, using 4 corrosive links. These corrosive links cause the ballast weight to be jettisoned after a period of 12 hours, thereby improving the prospects of successful recovery in the event of a vehicle system failure. Although theoretically buoyant for seawater, the vehicle proved to be correctly ballasted without any additional weight. Salinity tests confirmed the harbor water to be slightly brackish at 26-28 parts per thousand, thus helping to account for the density disparity. A small ~300 cc piece of floatation foam was added inside the bow of the vehicle to achieve neutral trim.

Kemonaut and NEREUS clock synchronization and systems checks were performed subsequent to ballasting. After satisfying all checks, the AUV was released from its lift hoist and towed into position using a 16 ft skiff equipped with a 35 hp outboard engine. Once in position, mission commands were relayed to Kemonaut from the base station using the spread spectrum radio modem and operations commenced once the vehicle was in position.

Figure 6: Nautical Chart of Boston Harbor
showing GIB placement and overall area of operation
view an enlargement

Two initial surface missions were conducted to verify proper vehicle tracking with the GIB hydrophone array, while maintaining visual contact. Following the two surface runs, Kemonaut performed a total of 6 submerged missions, ranging in duration from 2 to 6 minutes and reaching a maximum depth of 6.3 meters. Each dive mission was conducted as level, straight line, run for a specified time duration with the aim of reducing mission complexity and keeping AHRS error from critically affecting the navigation system. The navigation system was set to abort its mission if the vehicle's sonar altimeter sensed a distance of less than 3 meters from the bottom to prevent the vehicle from colliding with the harbor bottom. The intercept boat tracked the submerged AUV using GIB hydrophone position fixes of the vehicle relayed verbally over VHF radio from the base station. After each dive the intercept boat was sent to visually inspect Kemonaut for obvious signs of damage and to prevent other nearby vessels from colliding with the somewhat concealed AUV. In addition to visual inspection after surfacing, NEREUS instrument status was evaluated via a dedicated radio link from a laptop computer located in the intercept boat. As a measure of caution, missions were conducted verify proper vehicle tracking with the GIB hydrophone array, while maintaining visual contact. Following the two surface runs, Kemonaut performed a total of 6 submerged missions, ranging in duration from 2 to 6 minutes and reaching a maximum depth of 6.3 meters. Each dive mission was conducted as level, straight line, run for a specified time duration with the aim of reducing mission complexity and keeping AHRS error from critically affecting the navigation system. The navigation system was set to abort its mission if the vehicle's sonar altimeter sensed a distance of less than 3 meters from the bottom to prevent the vehicle from colliding with the harbor bottom. The intercept boat tracked the submerged AUV using GIB hydrophone position fixes of the vehicle relayed verbally over VHF radio from the base station. After each dive the intercept boat was sent to visually inspect Kemonaut for obvious signs of damage and to prevent other nearby vessels from colliding with the somewhat concealed AUV. In addition to visual inspection after surfacing, NEREUS instrument status was evaluated via a dedicated radio link from a laptop computer located in the intercept boat. As a measure of caution, missions were conducted at progressively deeper target depths of 2, 4 and 6 meters. Mission duration was also increased incrementally from 2 minutes on first dive to a maximum of 8 minutes on the sixth dive.

Figure 7: Kemonaut pressure transducer record

After satisfactory completion of its first and second dive to respective depths of 2 and 4 meters, Kemonaut was requested to dive to a depth of 6 meters on its third mission and reached a maximum depth of 6.3 meters before resurfacing. At the time it was unclear if the bottom avoidance override had been triggered by an acoustic return from an obstruction lying on the harbor bottom. Although later analysis of sonar altimeter data revealed that this had not been the case, in the interest of vehicle safety, a decision was made at that time to conduct the next dive at a depth of 4 meters, but with an increased mission duration of 5 minutes. Kemonaut executed this dive mission cleanly by quickly rotating itself nearly 180° to assume the proper heading and then maintaining its straight course for the remainder of the mission, logging some 400 meters in distance. The fifth dive mission was setup to better ascertain the repeatability of the vehicle's navigation and control system by retracing this fourth dive. Consequently, Kemonaut was towed back to the vicinity of its previous starting position and instructed to dive to a depth of 5 meters for a duration of six minutes using the same heading as dive 4. Again, Kemonaut performed capably, following to within 1 meter of its previous track line for a distance of over 200 meters. The sixth and final dive mission was intended to further test the AUV's ability to maintain a straight and level track line and to allow the NEREUS instrument adequate time to collect multiple chemical spectra while at depth. Mission duration was specified for 8 minutes and a shallow depth of 3 meters was chosen in order to minimize the chances of bottom collision. The desired heading was increased in a similar attempt to prevent the AUV from being run aground on the ruins near the harbor channel's southern sea wall and Coast Guard piers. Despite these efforts, Kemonaut was forced to abort the dive mission approximately 4 minutes into the run because the assigned heading had inadvertently overcompensated for the current, causing the vehicle to swim northward, passing underneath the USS Cassin Young and a barge before swimming into a shallow dry dock area, causing the bottom avoidance override procedure to trigger, thus forcing the vehicle to surface.

Based on recorded MOOS pressure sensor data, Kemonaut has demonstrated a dive rate capability of up to 0.18m/s, and is able to maintain a prescribed depth to within ±0.3 meters (Figure 7). Despite uncertainties caused by variability in near-surface acoustic signal propagation, GIB hydrophone tracking revealed that Kemonaut was able to maintain a high degree of lateral accuracy (Figures 8 and 9). The extent of Kemonaut's dynamic control is evident when comparing the GIB track logs from dive missions four and five, particularly in light of the fact that these dives were performed while in the presence of a current with a flow rate exceeding 1.5 kts.

Figure 8: 2-dimensional GIB hydrophone surface track log of Kemonaut dive missions

Figure 9: 3-dimensional GIB hydrophone track log of Kemonaut dive missions

The NEREUS instrument, which operated for the duration of its time aboard the Kemonaut vehicle, collected approximately 90 individual spectra over the course of these three hours, with each spectrum consisting of 693,000 discrete measurements. The NEREUS instrument appeared to be largely unaffected by vehicle motion through the water or vibration from the vehicle's propulsion and control system while underway.

The NEREUS instrument conducted spectral scans of volatile dissolved gases ranging in molecular weight from 12 to 150 Daltons while operating onboard Kemonaut. These scans were performed at a mass step interval of 0.1 Dalton to ensure that ion peak shape could be clearly resolved. Signal averaging was kept to 500 samples per data point in an effort to minimize the amount of time required for a complete spectral scan while still maintaining an adequate S/N ratio. Automated radio communications between the instrument and its remote computer terminal (used for monitoring the instrument) were scheduled to occur during the electrometer stabilization period before the instrument took a measurement to further decrease time requirements. NEREUS was able to complete an entire scan once every 98 seconds using these techniques. Although spectral scans can be conducted much faster by further limiting the mass-to-charge range, mass step interval and signal averaging, this mode was chosen so that high molecular weight hydrocarbons could be clearly identified in the event of being encountered by the AUV.

Post-deployment analysis of NEREUS data revealed the presence of dissolved atmospheric gases, principally nitrogen, oxygen, and argon (Figure 10). Due to the cold temperatures encountered on this day (an air temperature of -5 to -1° C and a water temperature of 2 to 4° C) the instrument's inlet membrane permeability was significantly decreased, causing all spectra to exhibit attenuated peak heights. Nevertheless, relative ratios of these gases remained consistent with previous NEREUS data collected during buoyed deployment in this area of Boston Harbor. RMS noise was typically about 5 mV, but exhibited sudden dramatic increases, to hundreds of millivolts, when the intercept boat occasionally bumped into the AUV while it was being towed.

Figure 10: A typical NEREUS spectrum that was recorded while onboard the Kemonaut AUV, showing dissolved nitrogen, oxygen and argon.

NEREUS spectrum data was processed upon completion of the six dive missions to quantify ion peaks and then merged with MOOS track log data by way of time stamp synchronization. The resulting data shows a noticeable variability in dissolved gas concentration near the water’s surface. This anomaly is likely to be the result of air bubble entrainment from breaking waves generated by the +20kt wind blowing that day. Further analysis of the data did not reveal the presence of any hydrocarbon fractions or xenobiotic molecules.

Figure 11: Dissolved gas ion signal as a function of depth

Future improvements may include upgrading the vehicle power supply to a high-energy density secondary battery system to allow for extended missions and quick recharge. Implementation of a closed loop navigation system could greatly expand the vehicle’s ability to operate submerged for longer periods and to undertake surveys where more precise positioning is required. Finally, integration of the NEREUS adaptive control system to the Kemonaut main vehicle computer may enable the system to intelligently conduct chemical surveys without a priori knowledge of the survey area.

• Bellingham, J. G. (1997). "New oceanographic uses of Autonomous Underwater Vehicles." Marine Technology Society Journal 31(3): 34-47.
• Bellingham, J. G., C. A. Goudey, et al. (1994). A Second Generation Survey AUV. Proceedings of the IEEE Symposium on Autonomous Underwater VehicleTechnology (AUV '94), Cambridge, Ma., IEEE.
• Damus, R., S. Desset, and J Morash (2002). Xanthos vehicle control system. R. Camilli.
• Hemond, H. and R. Camilli (2002). "NEREUS: engineering concept for an underwater mass spectrometer." Trac-Trends in Analytical Chemistry 21(8): 526-533.
• Manley, J. and J. Rieffel (2000). GOATS 2000 September 25-October 12, 2000 AUV Operations Report, MIT Sea Grant: 13.