News
About us
Products
3D Imaging and Mapping
Custom LADAR Systems
Down-Hole 3D Imager
Mobile Underwater 3D Wall Mapper
Sonar Mapping and Obstacle Avoidance System
Autonomous Vehicles and Systems
AUV Propulsion Controller
DEPTHX
ENDURANCE
Guidance Navigation and Control Systems
Leak Detector System for AUVs
Unmanned Vehicle System Executive
Variable Buoyancy Engine
Closed-Cycle Life Support (ECLSS and PLSS)
Cis-Lunar MK5 PLSS
Head Up Displays for PLSS Units
Manned Closed Cycle O2 Control Systems
Manned ECLSS CO2 Processors
Custom LCD Displays and Firmware
Manned and Unmanned Propulsion Systems
Mobile and Portable Power Systems
High Density Vehicle Power Systems
Lightweight Portable Li-Ion Power Systems
Through-Rock Communications and Localization Systems
Smart Devices
Contact

 

 

Home | Contact

ENDURANCE
(Environmentally Non-Disturbing Under-ice Robotic ANtarctiC Explorer)

ENDURANCE is the Stone Aerospace Phase 2 follow-on design to DEPTHX. Designed to test down-hole (through ice) deployment of an autonomous “fast mover” AUV that characterizes large sub-surface volumes -- while not influencing sensitive biological and chemo/thermo stratifications -- and reliably auto-docks with an extraction system at the melt hole at the conclusion of a mission. The research work is targeted to be conducted in Lake Bonney, Antarctica, in 2008-2009.

ENDURANCE Mission Description:

The principal AUV (Autonomous Underwater Vehicle) mission at Lake Bonney is to serve as a tether-less programmable mobile platform for an onboard suite of science probes. The concept (upper right figure) is to deploy a compact vehicle through a melt hole in the ice surface of the lake and thence to have the vehicle execute a pre-programmed 3D grid sampling trajectory in order to populate a voxel model of the scan volume with the scalar science data. Each layer of the grid is scanned in strips, raster style, with a strip width, X, and length, L. In a square search pattern samples will be taken in longitudinal slices of width X and the length of any side would be n*X, where n is the number of scan lines. This can be taken a step further by rastering in the depth direction by an increment Z between layers. A sample volume equal to X2Z would represent 1 voxel. The present grid being considered for the ENDURANCE project is 10 m (width) x 10 m (length) x 1 m (depth) for fine resolution sampling. The grid is geo-referenced to the melt hole coordinates (using GPS), and the sub-ice positions are determined using a multi-stage sensor fusion approach. Each geo-referenced voxel is then populated with fiber sensor data (fluorescence spectrometer; Raman spectrometer; dissolved oxygen; temperature; ambient light; Raman bottom probe); aqueous chemistry (chloride and conductivity); and an optical image. Data for any given mission are stored redundantly on flash disk for later recovery and analysis on the surface. Planned mission duration is 8 hours between vehicle battery changes and maintenance and it is planned for at least one mission per day to be conducted in the field. The navigation system is designed to permit the vehicle to work autonomously. However, for safety and recovery surety reasons, a backup nav system will be used that will limit range of up to 1,000 m from the melt hole.

Vehicle Description

Because ENDURANCE must be deployed “down hole” a significant design effort is required to develop the most compact vehicle possible to enable ingress and egress through the smallest possible melt hole in the lake surface. The nominal dimensions for the compact vehicle are 1422 mm in length, 1066 mm in width, and 782 mm in height with an estimated dry weight of 80 kg in air and neutral and non-rotating in water. Other architectures are possible and are being considered for the final design.

In an effort to deliver an extremely capable vehicle at minimum cost the vast majority of the components are COTS. The propulsion chassis uses commercial ROV components along with custom-designed flight electronics and thrust vector controllers. Maximum vehicle speed is anticipated to be between 1 to 1.5 m/s, sustainable for a design mission duration of 8 hours daily. Onboard power is a major concern in any AUV application. Meeting the 8 hour mission duration spec in a compact, moderately lightweight vehicle – in order to make it “man-transportable” in Antarctica – forces the use of sub-sea rated high density power technology.

The majority of the science payload sensors are fiber-optic based and will be mounted as a series of inline devices along the wall of a flow-thru tube that runs the entire interior length of the vehicle fed by a bow intake port. Chloride and conductivity probes will sense the same flow stream at the aft end of the vehicle. A CCD-based digital camera, with lighting, is located on the starboard bow for capturing visible spectrum images within each voxel as well as capturing bottom sediment images.

Navigational Architecture

The heart of any AUV is its ability to know its location in three dimensional space without the use of human guidance (tethered or otherwise). The figure at right illustrates the navigational design being developed for ENDURANCE. The auto-docking system employs a spread-spectrum USBL (Ultra-Short BaseLine) transceiver down-hole at a depth equal to half the maximum water depth at the melt hole. The ENDURANCE vehicle carries a transponder which allows for a full-duplex data comm link. At Lake Bonney the vehicle transponder will send out an encoded “ping” that is detected by the down-hole transceiver. The received signal is then processed and a slant range and direction (relative to a geo-referenced coordinate system established at the down-hole) are deduced. This information is then encoded and returned to the vehicle via a spread spectrum burst from the transceiver. The USBL will be the primary auto-docking navigation tool on ENDURANCE. But it is not sufficient for precision autonomous operation, and abort-scenario recovery must also be provided in the event of failure of the USBL during a mission. ENDURANCE determines its depth below the water surface using a composite sensor system described below. Attitude (and in particular Yaw) is determined by a high resolution FOG (fiber optic gyro IMU). This latter instrument, when combined with a 3-axis Doppler sonar array (which determines absolute velocity components in three orthogonal directions) provides a completely independent method of navigating the vehicle in the absence of the USBL. The IMU attitude output is quite stable – with a drift of less than 0.1 degree per hour – but positional drift (owing to computational round-off and noise affecting the double integration of accelerometer data) can be significant. The Doppler sonar compensates for this by providing absolute velocity estimates. All of the above – USBL, IMU, Doppler sonar, and depth measurements – will be fed into a real-time mathematical filter that will, at any given time, provide the vehicle with a best estimate of position and orientation relative to the down hole. Abort recovery procedures are discussed below. Using this approach we presently estimate peak location errors of less than 5 meters horizontally and 0.5 meter vertically at 1,000 m range from the down hole, well within the desired high resolution voxel size of 10 m x 10 m x 1 m.

Depth Compensation

Data from prior single-point (down-hole) observations at Lake Bonney and other Antarctic lakes have shown a relatively striking change in water characteristics from the ice roof to the lake bottom, with salinity varying from that of fresh water in contact with the roof to 4 times that of seawater at the bottom. These changes are reflected in vertical stratification of density, with the density increasing as one proceeds towards the bottom. Varying density affects the accuracy of both pressure transducer-derived depth estimates, and, to a lesser extent, vertical sonar range measurements to either the ice surface or the lake bottom due to the fact that acoustic wave propagation velocity is proportional to the square root of fluid density. Real-time measurements on the vehicle of chloride and conductivity (both measures of salinity and hence density), in combination with pressure transducers and averaged simultaneous sonar readings to the ice surface and lake bottom, will form the basis for a depth compensation algorithm that will run in real-time on ENDURANCE to provide the best estimate of true depth within any given voxel. Calibration studies will verify the validity, and error bounds, of the algorithm prior to deployment in Antarctica.

Obstacle Detection and Avoidance

Aside from navigation the second most important task running in real-time on the AUV is the detection and avoidance of obstacles. ENDURANCE will handle this through a series of bow, port, and starboard sonar arrays that create a 3D envelope of what exists within a 100 m radius of the vehicle. A total of 24 individual beams will be available that will permit updating of the perimeter world map at 12 Hz rates, leaving plenty of maneuvering time (since it will take the vehicle at least a minute to reach features detected at the far range of the sonars. In reality, the most likely significant feature variations will take place at the ice roof where pressure ridges may have formed and cause step-wise changes in the surface. Such features will be of limited depth extent and the establishment of a minimum “keep out” depth will insure free running of the vehicle.

Minimization of Layer Mixing and Bottom Disturbance

In order to prevent turbulent mixing of thermoclines, haloclines, and other important chemo and density gradients we intend to conduct CFD (computational fluid dynamics) analyses of ENDURANCE in a mixed layer fluid that models the anticipated layering to be expected at Lake Bonney and the other field sites. These pre-design studies will confirm maximum acceptable vehicle velocities, the impact of proposed vehicle geometries on turbulent wake generation, and the acceptable closest-point-of approach to the lake bottom to avoid sediment stirring.

Using the redundant absolute depth and lake bottom range from the vehicle it will be possible to reliably insure that the vehicle never approaches close enough to the lake bottom to cause turbulence that would stir the sediment layering. However, there is strong interest on the part of the science team to obtain Raman spectroscopy samples of this sediment. The solution designed for ENDURANCE is a servo-operated fiber optic spooler that lowers a small diameter fiber optic Raman sensor into the sediment from a height of several meters off the floor and then retracts it before moving to a new sampling position.

Mission Abort Contingency

Loss of the vehicle, and its subsequent decay on the lake floor is considered an unacceptable scenario for this field mission. We have taken this design criteria into account at three separate levels. First, the USBL spread spectrum tracking system will report the location of the vehicle in real-time to the surface-based Mission Planner. It will be possible to customize the downlink to the vehicle to include a message that can be interpreted by the onboard mission manager (OMM) as a remote mission override and subsequently to either cause the vehicle take situation-specific recovery measures.

Two independently-powered subsystems will activate if an abort-to-surface directive is received by the OMM or if the vehicle suffers a full loss of main power or a computer hardware or software crash. The first such system is a servo-pneumatically activated positive buoyancy device that will cause the vehicle to rise until it encounters the ice roof. Simultaneously, a magnetic induction transponder (MIT) is activated. The MIT will generate an AC magnetic field at a frequency and power level that will penetrate up to 1200m of ice and rock. The central axis of the field is known as the “null” axis and along it the magnetic field strength effectively vanishes. Recovery personnel on the surface will be able to use phase-lock loop directional antennas to precisely locate the surface penetration point of the field axis. A new melt hole will then be created at that location for subsequent vehicle recovery. The accuracy of the Stone Aerospace system has been proven in much more difficult conditions (through rock) to be less than +/- 0.1 m on clear days and about +/- 1 m when thunderstorms are present.

Contact our engineering staff for further information.