Cornell University Autonomous Underwater Vehicle: Design and Implementation … · 2019. 10. 23. · Abstract—CUAUV Argo is the new 2014-2015 autonomous underwater vehicle (AUV)

Cornell University Autonomous UnderwaterVehicle: Design and Implementation of the

Argo AUVNicholas Akrawi, Mustafa Ansari, Linda Barron, John Bartlett, Zander Bolgar, Page Bowers, Jonathan Chan,

Corey Chang, Eric Chiang, Nancy Ding, Kent Esslinger, Joseph Featherston, Christopher Goes, Andrew Halpern,Melissa Hamada, Daniel Harianja, Yaseen Islam, Brendon Jackson, Dong-Hyun Kim, Moonyoung Lee (co-lead),

Noah Levy, Victor Lucena, Michael Mahoney, Alexandru Malcoci (co-lead), Mihir Marathe, Artina Maloki,Michael Ngyuen, Zachary Porter, Alexander Renda, Charles Sanderson, Daryl Sew, Alexander Spitzer, Ricardo

Stephen, Emily Sun, Ellen Thiel, Ian Thompson, Nancy Wang, James Wu, Shuqing Wu, and Mei Zhang.

Abstract—CUAUV Argo is the new 2014-2015autonomous underwater vehicle (AUV) designed andbuilt by a team of 40 undergraduate students atCornell University. Completed in a ten month de-sign cycle, the vehicle was fully modeled usingCAD software, extensively simulated with ANSYS,and manufactured almost entirely in-house. With15 years of aggregate research and development toenhance autonomous underwater vehicle technologyby CUAUV students, this year’s vehicle is a furtherimprovement on all previous designs. Argo presentsa stronger, lighter, and more agile platform than everbefore.

Particularly notable new advancements includecustom brushless thrusters - two of which aremounted on stepper motor driven vector modules,a full PCB plug-and-play backplane, an upgradedPython-based vision system driving new state-of-the-art cameras, and a refactored mission system allow-ing for superior task execution control. Argo’s sensorsuite includes compasses, inertial measurement units(IMUs), a Teledyne RDI Explorer Doppler VelocityLog, a depth sensor, an internal pressure sensor, ahydrophone array, a BlueView high-definition imag-ing sonar, and both forward and downward-orientedCMOS cameras. Returning features include a dual-cantilevered electronics rack, a vacuum-assisted seal,hot-swappable batteries, pneumatic actuators, andunified serial communications.

I. INTRODUCTION

THE Cornell University Autonomous Un-derwater Vehicle team’s primary objective

is to design and build an autonomous underwa-ter vehicle (AUV). The AUV is a cumulativeproduct of many integral end-to-end projectsthat necessitate extensive hands-on experience,critical problem solving skills, and interdisci-plinary collaboration amongst students. Uponcompleting the vehicle in a ten month designcycle, CUAUV participates in the annual AU-VSI Foundation and ONR International Ro-boSub Competition. The competition is heldin late July at the TRANSDEC facility, partof SPAWAR Systems Center Pacific in SanDiego, California. The competition is designedto challenge student-built AUVs with an obsta-cle course that simulates real-world AUV mis-sions such as precise mapping of the seafloorand construction of subsea infrastructure. Ro-boSub competition tasks range from shape andcolor recognition to torpedo firing and finemanipulation of small objects. Each of thesemissions must be completed by the vehicleindependently (without human interaction orcontrol). Successfully completing such tasks re-quires accurate visual and acoustic detection ofcompetition elements, fine-control navigation,obstacle avoidance, and object manipulation. Inorder to complete the task of building a vehiclethat is capable of navigating all mission ele-ments and meeting all requirements, CUAUVis divided into mechanical, electrical, softwareand business/public-relations subteams.

Cornell University Autonomous Underwater Vehicle 1

II. DESIGN OVERVIEW

The 2014-2015 primary vehicle, Argo, is ahovering, littoral-class AUV designed primarilyto compete in the RoboSub competition. Thedesign is similar to that of work or observa-tion class remotely operated vehicles (ROVs),primarily targeting space-constrained complexfunctionality and fine-grained positional con-trol.

Fig. 1: A SolidWorks rendering of CUAUV’s2015 vehicle, Argo

Argo is an improvement over CUAUV’sprevious vehicles in terms of fine control, ro-bustness, and ease of assembly. The robustnesswas improved through more extensive electricaltesting, rigorous finite element analysis (FEA),and increased usage of computer numericalcontrol (CNC) machining. Argo features eightthrusters which give it control over six degreesof freedom and improved stability in move-ment. The vehicle also includes a pneumaticactuator system that allows it to interact withits environment. Argo measures 46 inches inlength, 27 inches in width, and 25 inches inheight. Its dry weight is 80 pounds.

Custom electronic boards designed by stu-dents are isolated in the aft hull and work asinterchangeable cards connecting to a centralPCB backplane via card edge connectors. Thevehicle is powered by two lithium-polymerbatteries and contains modular power, sensor,and serial communication systems. A full on-board suite of visual, acoustic, inertial, andpressure sensors is used for navigation and datacollection. The vehicle’s software suite is run

by a single-board computer with a quad-corehyperthreaded Intel i7 CPU, and is primarilycomposed of intercommunicating shared mem-ory, serial, control, vision, and mission systems.

III. MECHANICAL SYSTEMS

Argo’s mechanical system consists of thevehicle frame, upper hull, actuators, and ex-ternal enclosures. The upper hull and externalenclosures are responsible for protecting theelectronic components from water, while thestructure provides mounting points and pro-tection for sensors, enclosures, and actuators.Assemblies are designed in SolidWorks, sim-ulated using ANSYS FEA software, and thenmanufactured either manually or with the helpof CNC machinery and computer-aided manu-facturing (CAM) software (Pro/TOOLMAKERand FeatureCAM).

A. Frame

The frame defines the positions and orien-tations of each mechanical component in thevehicle, maintaining the structural integrity andrigidity of the vehicle and protecting delicatecomponents. This year’s frame implements amodular method of component attachement.Waterjet cutting also allowed for rapid man-ufacture of frame components, allowing forspeedy assembly and testing.

Fig. 2: A SolidWorks rendering of Argo’sframe.


Most of the components on the vehicle arescrewed directly to the frame, requiring noextra mounting features. This saves weight,reduces complexity, and allows for efficient useof space, All components are placed such thatthe center of mass is at the relative center ofthe AUV. The frame emphasizes ease of use,manufacturing, and integration whilst remain-ing adaptable.

B. Upper Hull and Electronics Rack

Argo’s upper hull is composed of two racksjoined by a midcap, undersea connectors, theDVL transducer head, and two acrylic hullassemblies that protect the electronics. Thefore rack contains all commercial off-the-shelf(COTS) components in the upper hull includingour forward camera, COM Express module,DVL electronics, and LCD display. The aft rackcontains all custom electronics linked togethervia an easily customizable backplane system.All boards slide into connectors via 3D printedrails, and the rack also raises to a vertical posi-tion, allowing for simple and fast maintenance.

Fig. 3: A SolidWorks rendering of Argo’s foreand aft racks.

All connections to the outside of the upperhull are made via SEACON connectors locatedon the midcap. The midcap provides hingedmounting for both the fore and aft racks, theDVL transducer head, and the sensor boom.Although difficult to manufacture, the midcapmakes efficient use of space and material,while minimizing the risk of leaks. The twoacrylic hulls slide over the cantilevered racksto provide sealing. The fore hull has a cameraenclosure built in that slides over the protruding

camera. Redundant sealing of the hulls as wellas a slight vacuum within the upper hullsrenders hull leak risk negligible.

C. ActuatorsThe actuators system is comprised of two

torpedo launchers, two marker droppers, twoactive grabbers, and one passive forward ma-nipulator. A 3000 psi paintball air tank regu-lated down to 100 psi serves as the air supplyfor the pneumatic portion of the system. Twelveinternal and two external valves, the formerhoused inside an integrated manifold and valveenclosure, pass air from the valves to the restof the pneumatic actuators system. With thisconfiguration, Argo can independently fire in-dividual torpedoes and markers, actively graband release two objects simultaneously, andeffectively manipulate forward elements.

1) Torpedo Launcher and Marker Drop-per: The torpedoes and markers are customdesigned and cast projectiles made to travelaccurately through the water. Torpedoes arepropelled pneumatically and have a range ofapproximately 15 feet. Markers are held inplace by magnets until a small air burst unseatsthem, after which their path downwards is heldstraight by large fins and heavy tips.

Fig. 4: One of Argo’s downward grabbers


2) Active Grabbers: Two active grabbers areresponsible for grabbing the recovery objects.Each grabber consists of a pneumatic linearcylinder connected to a claw on an actuated ex-tending arm. The cylinders are double-action;firing one valve closes the claw and firing theother opens it.

3) Forward Manipulator: Argo features apassive forward manipulator utilized to remove”lids” on mission elements. An X-shaped con-tact element serves to automatically guide thehandle of the lid into the sub’s grasp.

4) Thrusters: Propulsion is provided byeight custom-machined brushless thrusters.Each thruster has a built-in control board al-lowing for position-independent hot swapping.The two aft thrusters are mounted on stepper-controlled mounts, enabling dynamically di-rected thrust vectoring. Argo’s thruster mount-ing scheme provides the vehicle with activecontrol in all six degrees of freedom.

D. External Enclosures

Argo’s external enclosures contain variouselectrical systems located outside of the upperhull pressure vessel. These waterproof vesselsisolate the systems from unwanted noise andallow for flexibility in the construction of thevehicle. Systems contained in external enclo-sures include the hydrophone passive acousticsystem, kill switch, sensor boom, downwardcamera, and battery pods.

1) Hydrophones Enclosure: The hydro-phones system is kept separate due to noiseconsiderations. The enclosure was designed tobe as light as possible, and is constructedentirely out of aluminum to facilitate effectivecooling. In addition, the piezoelectric elementsof the hydrophones system are mounted di-rectly to the wall of the enclosure, eliminatingthe need for a separate mount for the elementsand keeping the system compact.

2) Sensor Boom: Some of the vehicle’s sen-sors must be isolated from the electromagneticnoise caused by thrusters and other high powerelectrical components. The sensor boom is

mounted atop the midcap far from any othernoise sources in order to maintain signal in-tegrity. It contains a LORD MicroStrain 3DM-GX4-25 AHRS, and a custom team designedcompass and IMU. Signals are passed back tothe upper hull via SEACON connectors.

3) Battery Pods: Each battery pod featurestwo SEACON connectors, one for monitoringtemperature, and the other for charging anddischarging the batteries. A DeepSea pressurerelief valve also prevents any buildup of inter-nal pressure, removing the danger of explosiondue to outgassing. The battery pod lids arelatched to the hull for quick battery removaland replacement.

Fig. 5: Argo’s battery pod enclosure

IV. ELECTRICAL SYSTEMSArgo’s electrical systems supply up to two

hours of reliable power throughout the vehi-cle and provide an interface between the on-board computer and all sensors and peripheraldevices. The electrical subteam designs andpopulated custom circuit boards with the assis-tance of various schematic and PCB softwarepackages. The dual-hull design of the vehicleemphasizes the strategic placement of electron-ics and wire management by dividing COTSdevices from custom electronic boards.

As a result of this and improved PCB back-plane usage, electrical connections are opti-mized between boards inside the hull, externalenclosures, and sensors exterior to the upperhull. The amount of time required to maintainthe electrical system is decidedly reduced rel-ative to previous designs.


Fig. 6: Layout schematic of Argo’s brushlessmotor control board

A. Power System

The power to run Argo is provided by twoAdvance Energy 8000 mAh lithium polymerbatteries, which give the vehicle a run time ofapproximately two hours. In order to maintainconstant uptime while testing, Argo utilizeshot-swappable batteries, allowing for quick bat-tery changes that do not require a full systemreboot. To facilitate use of the vehicle on shore,a bench power station has been developed topower the sub from a standard AC powersource. The input power sourced from the twobatteries is routed through the pod board, whichcombines the two power sources to provide asingle power rail for the vehicle. To ensurenominal and equal discharge from both batterypacks, the pod board constantly compares thevoltages and drains power from the battery withhigher potential.

In order to ensure reliably regulated power,the power distribution board provides the elec-trical system with isolated power rails at +5,+12, and +24 Volts. The board measures poweruse from each port and passes these statis-tics to the computer. In the case of excessivecurrent draw, the computer has the ability toshut down the corresponding port to prevent

Fig. 7: Block diagram of Argo’s power infras-tructure

potential damage to the components.

B. Serial Communication

The serial board is the interface between thecomputer and the various sensors and custom-designed boards. It allows twelve devices tocommunicate via RS-232 serial through a sin-gle USB connection. The RS-232 protocol waschosen because of noise tolerance and ease ofuse and deployment.

C. CAN Bus

The CAN protocol is used to communicatewith each of the eight thruster modules aswell as the two vector thrust modules. CANwas chosen over RS-232 because all the con-nected devices share only two communicationwires. This allows the connection to all of thethrusters and the vector thrust modules to beidentical and allows connections to be swappedwithout effect. Each board is issued a uniquedevice ID to enable individual thruster control.

D. Thruster Control

Each of the eight thrusters contains a custombrushless motor control board. Each board con-tains a microcontroller to communicate with


the computer and control the motor. To abideby safety regulations, all thruster modules,vector thrust modules and the actuator boardcan be halted either by hardware switch orsoftware.

E. Vectorized Thruster Control

Each vector thrust module contains a steppermotor a custom controller board. This boardcontains two H-bridges to drive a bi-polarstepper motor and communicates with the com-puter on the CAN bus. The board shares aconnection interface with the brushless thrustermodules, simplifying connections and allowingthe devices to be swapped without softwareremapping.

F. Actuator Control

The actuator board control the solenoidsnecessary for all of the pneumatic manipulatorson the sub. The actuator board also acceptscommands to control the solenoids and anadditional two DC motors or one stepper mo-tor through communication with the computerusing an isolated serial line.

G. Sensor Interface

The General Purpose Input Output (GPIO)board has multiple ports to read and writeanalog and digital inputs to and from sensors.It is primarily designed to provide the electricalsystem with extra modularity, as there is oftena need to add sensors or electrical devices laterin the design process. In addition to readingmeasurements from the depth sensor, GPIOmonitors internal pressure of the vehicle toensure that partial vacuum is maintained.

V. SENSORS

In order to autonomously navigate throughthe competition course, the vehicle is equippedwith two main classes of sensors: sensorsto observe the vehicle’s external environment,and sensors to determine the vehicle’s internal

state. Visual recognition and navigation tasks inTRANSDEC are successfully identified usingone forward and one downward camera, whilethe recovery task is handled by the hydro-phones system using a passive acoustic array.The state of the vehicle is measured usinginertial measurement units (IMUs), compasses,a depth sensor, a Doppler Velocity Log (DVL),and a high-definition imaging sonar.

A. Hydrophone Array

Argo’s hydrophone system uses three Re-son TC-4013 piezoelectric elements to detectincoming acoustic waves. The design was op-timized for functionality and affordability uti-lizing a microcontroller and custom analogfiltering to interpret the element data. Thehydrophone system accurately calculates theheading and elevation of a pinger relative tothe vehicle within one degree. The hydrophonesystem has also been extended to stream dataover Ethernet to the vehicle’s main computerwhich hosts a platform for real time acousticdebugging and finer tuning of the overall sys-tem.

B. Orientation

A combination of IMUs and compasses mea-sure the vehicle’s acceleration, velocity, andspatial orientation. The orientation sensors onthe vehicle are a MicroStrain 3DM-GX4-25AHRS and a team-designed compass and IMU.An MSI Ultra-stable 300 pressure sensor mea-sures the depth of the vehicle.

C. Doppler Velocity Log

The Teledyne RDI Explorer Doppler Ve-locity Log (DVL) provides accurate three-dimensional velocity data. This information isused in conjunction with the other sensors toprovide closed-loop vehicle control.


VI. SOFTWAREAll of Argo’s higher level functionality,

including autonomous completion of missiontasks, is achieved through the vehicle’s soft-ware system. The software stack is built uponthe Debian GNU/Linux operating system andincludes a custom shared memory state syn-chronization system, a serial daemon for com-munication with the majority of the electri-cal boards, multithreaded vision, model-basedcontrol, and abstract task/mission systems. Ourcustom software is primarily written in C++and Python, although certain individual subsys-tems are written in Go and Haskell.

Fig. 8: Argo’s software stack

A. ComputerThe software on the vehicle is powered by

an Intel Core i7-4700 Haswell quad core pro-cessor on a Connect Tech COM Express type 6carrier board along with an ADLINK Express-HL module, using a 256GB mSATA solidstate drive (SSD). The computer is connecteddockside through a SEACON Gigabit Ethernettether.

B. Shared MemoryThe shared memory system (SHM) provides

a centralized interface for communicating thestate of the vehicle between all running pro-cesses, electronic subsystems, and users con-trolling the vehicle. SHM is a custom system

built upon POSIX shared memory, providingthread- and process-safe variable updates andnotifications. Various elements of the vehiclestate are stored in and read from shared mem-ory. These shared variables can be accessed byall of the components of the software systemconcurrently, which allows for simple commu-nication among the various daemons.

C. Unified Serial Daemon

The Unified Serial Daemon (USD) handlescommunication between Argo’s on-board com-puter and the internal electrical boards by im-plementing a standardized serial protocol. Withthe USD, a single configurable daemon on thecomputer is able to communicate with all ofArgo’s custom serial-capable boards. A varietyof functionality is built into the USD protocol,such as board identification and microcontrollermonitoring.

D. Unscented Kalman Filter

In the past, CUAUV has made use of stan-dard Kalman filtering in order to process thelarge amounts of raw data from the AUV’sextensive sensor suite. Utilizing Bayesian tech-niques in conjunction with raw sensor data topredict future states of the vehicles positionand orientation, Kalman filtering is able to effi-ciently smooth erratic readings and compensatefor any unreliability in sensor data. Althoughthese standard filters have been effective in thepast, they perform sub-optimally when filteringnonlinear systems. Because of this, this year’svehicle features an Unscented Kalman filterwhich processes orientation data, in conjunc-tion with standard Kalman filter for position.By using statistical methods and the UnscentedTransform, this new filter is able to heavilyreduce the complexity of processing nonlinearsystems, while still generating accurately fil-tered output. The use of the Unscented filterallows us to use highly nonlinear quaternions torepresent orientation, which do not suffer fromthe same limitations as traditional Euler angles


(eg. Gimbal locking). This allows for a muchmore robust controller and overall enhancedstability for the vehicle.

E. Vision

Our vision-processing system is designed togive the vehicle up-to-date and accurate dataabout the surrounding mission elements. Adaemon for each camera captures and passesraw camera frame data through an undistortionfilter. The camera data is then passed intoshared memory for concurrent processing byeach vision module. The camera daemons andvision modules are entirely modular, so that wecan run any combination of modules on eitherlive data or replayed logs. The undistorteddata is then analyzed using a combinationof color thresholding, Canny edge detection,contour analysis, and Hu moment characteriza-tion among other methods. Argo supports oneforward-facing XIMEA xiQ USB3 camera witha wide-angle lens, and a downward-facing IDS-uEye-6230SE-C-HQ Gigabit ethernet camerafor the mission elements located below Argo.

Fig. 9: Vision system graphical user interface

1) Vision Tuning: A vision tuning systemenables fast, real-time adjustments to the visionparameters. The vision tuning user interfaceshows the results of intermediate steps andmakes it easy to see the outcomes of changes

as they are made, allowing us to rapidly iter-ate upon our vision analysis algorithms. Eachmission element has its own independent visionmodule. This allows for multiple modules to berun in parallel.

2) CAVE: The CUAUV Automated VisionEvaluator (CAVE) helps analyze vision per-formance by keeping a database of loggedvideo and providing a graphical framework forquick tag-based annotation and automated test-ing. CAVE organizes captured logs and allowssearching by metadata, including informationsuch as the weather conditions, location, andmission elements present in a video. Addition-ally, the system can play video files on demandand stream frames directly to the existing vi-sion framework, allowing for effortless testing.

Fig. 10: A screen shot of CAVE running onlegacy log data

F. Vehicle Logging and SimulationThe logging system captures the full shared-

memory of the vehicle during mission runs.Furthermore, a log playback utility allows thesoftware team to simulate the vehicle with realmission data and perform additional debuggingand development out of the water. This systemhelps isolate bugs which only occur rarely.

We additionally use a fast and extensible net-worked simulator to rapidly test mission logic.Work to extend the simulator with automatedmission unit testing and fuzzing functionality isin progress. The simulator interfaces with ourexisting shared memory subsystem, so simu-lated missions can be run without modification.


G. Mission System

A newly revamped mission system allowsfor direct control of code execution withoutsacrificing ease of reasoning. Tasks are createdin a simple functional style and can be com-posed to create various new useful tasks. Thesimplest combinators used are “Sequential” and“Concurrent” which run tasks in a serial orparallel fashion respectively. In order to con-struct robust mission tasks resilient to manyfailure cases, abstract inheritance and simplefinite state machines (“searching”, “centering”,“dropping”) are utilized to minimize the timeneeded to write and debug these additionalcomponents.

Fig. 11: ASLAM map of mission elements

H. ASLAM

In order to accurately navigate between var-ious unpredictably moving elements during arun, our sub constructs a dynamically generatedmap of objects within the pool, updating it asthe mission proceeds. The ASLAM (adaptivesimultaneous localization and mapping) suite,an extension and generalization of our previ-ous locator and layout algorithms, is built onstandard Bayesian statistics, using observationsof an object to probabilistically predict that ob-ject’s location. Instead of the covariance modelused last year, ASLAM constructs a probabilitydensity map for the sub in addition to all the

elements, serving to compensate for systematicpositional error. ASLAM additionally supportstime-based observational weighting, providingan increased level of responsiveness to non-static mission elements.

I. ControlUsing continuously collected and filtered ve-

hicle state data, Argo has precise and accuratevehicle control for all six degrees of freedom:surge, sway, heave, yaw, pitch, and roll. Thevehicle uses a empirically tuned PID controllerfor each degree of freedom along with a pas-sive force model and continuous optimization.The moment of inertia tensor of the vehicleis estimated from CAD to allow for bettermovement predictions. Using this system, Argocan compensate for any drag and buoyancyforces, even at non-trivial pitch and roll angles.The control system enables high-level missioncode to set desired values for the velocity,depth, heading, pitch, and roll.

VII. MISSION FEEDBACK ANDDIAGNOSTICS

Fig. 12: CUAUV’s LCD display

Argo features two bipartite ultra bright tri-channel RGB LED strips. The LED strips arepowered by a control board which takes inputfrom the vehicle’s software via RS-232. Theseprovide visual feedback of the vehicle’s currentstatus, which allows for continuous monitoring


of mission progress during untethered runs. TheLCD display is a new addition to this year’svehicle, and allows for quick and easy reportingof vehicle information without the need for acomputer. This is especially useful for systemdiagnostics, internal pressure monitoring, vehi-cle trimming display, and thruster testing.

VIII. VEHICLE STATUS AND TESTING

Argo is now in the pool testing phase. Priorto vehicle assembly, the mechanical systemswere thoroughly leak tested, and the electricalsystems were bench tested. The first in-watercontrol test took place in late spring. Testing isstill underway to prepare Argo for the RoboSubcompetition.

Fig. 13: A pool test of Argo in progress

ACKNOWLEDGMENT

CUAUV would like to thank all the indi-viduals who have supported us over the pastyear, including Cornell’s MAE, ECE, ORIE,and CS departments, and Cornell alumni whohave supplied our team with needed lab space,materials, tools, and funding. CUAUV wouldespecially like to thank our advisers: ProfessorsAlan Zehnder, Bruce Land, and Graeme Bailey;Director of Cornell Aquatics Fred Debruyn,the MAE Director of Instructional Laborato-ries Matt Ulinski, the Swanson Director ofEngineering Student Project Teams RebeccaMacdonald, Cornell Engineering’s administra-tive staff, especially Emily Tompkins, CarolMoss, and Maureen Letteer, and last but notleast the Cornell Rapid Prototyping Lab (RPL).

We would also like to thank all of ourcorporate sponsors, without whom we wouldnot be able to compete:

Alumni Sponsors: Erin Fischell; SamFladung; Kenneth Bongot; David Hayes; JohnDiamond; Greg Meess; Leila Zheng; JamesRajsky; Tom Brooks; Kirill Kalinichev; JamesMwaura; Brian Schiffer; Nick Elser; andDavid Hinkes.

Platinum Sponsors: Cornell University;Monster Tool Company; VideoRay;SolidWorks; Nexlogic; General Plastics;BlueView; Mathworks; SEACON; Shell; andAmerican Portwell Technology.

Gold Sponsors: Teledyne RD Instruments,LORD Microstrain, SeaBotix, Seaperch, Gen-eral Motors Foundation, Adlink TechnologyInc., and SGS Tool Company.

Silver Sponsors: CadSoft USA; Intel Cor-poration; LeddarTech; Molex; Advance EnergyInc.; iDS; ConocoPhillips; Lockheed Martin;PNI Sensor Corporation; Sparton; Allied Vi-sion Technologies; Surface Finish Technolo-gies; Exxon Mobil; General Electric; and Bit-tele.

Bronze Sponsors: Georgia Case Company;Adata; 3D Connexion; Sensata Technolo-gies; Digi-Key; Advanced Circuits; PolymerPlastics Corporation; The Mines Press Inc.;JetVend; Cayuga Xpress; MaxAmps; Pneuma-dyne; Domino’s Pizza; DIAB; Drexel Univer-sity; Asian Circuits; and Fast Circuits.

Fig. 14: The 2014-2015 CUAUV team withArgo


Documents

Cornell University Autonomous Underwater Vehicle: Design and Implementation … · 2019. 10. 23. · Abstract—CUAUV Argo is the new 2014-2015 autonomous underwater vehicle (AUV)