Stephen C. Martin

The Nereus hybrid underwater robotic vehicle for global ocean science operations to 11,000m depth

This paper reports an overview of the new Nereus hybrid underwater vehicle and summarizes the vehicles performance during its first sea trials in November 2007. Nereus is a novel operational underwater vehicle designed to perform scientific survey and sampling to the full depth of the ocean of 11,000 meters - almost twice the depth of any present-day operational vehicle. Nereus operates in two different modes. For broad area survey, the vehicle can operate untethered as an autonomous underwater vehicle (AUV) capable of exploring and mapping the sea floor with sonars and cameras. For close up imaging and sampling, Nereus can be converted at sea to operate as a tethered remotely operated vehicle (ROV). This paper reports the overall vehicle design and design elements including ceramic pressure housings and flotation spheres; manipulator and sampling system; light fiber optic tether; lighting and imaging; power and propulsion; navigation; vehicle dynamics and control; and acoustic communications.

Navigation and control of the Nereus hybrid underwater vehicle for global ocean science to 10,903 m depth: Preliminary results

This paper reports an overview of the navigation and control system design for the new Nereus hybrid underwater robotic vehicle (HROV). Vehicle performance during its first sea trials in November 2007 near Hawaii, and in May and June 2009 in the Challenger Deep of the Mariana Trench is reported. During the latter expedition, the vehicle successfully performed scientific observation and sampling operations at depths exceeding 10,903 m. The Nereus underwater vehicle is designed to perform scientific survey and sampling to the full depth of the ocean — significantly deeper than the depth capability of all other present-day operational vehicles. For comparison, the second deepest underwater vehicle currently operational worldwide can dive to 7,000 m maximum depth. Nereus operates in two different modes. For broad-area survey, the vehicle can operate untethered as an autonomous underwater vehicle (AUV) capable of exploring and mapping the sea floor with sonars and cameras. Nereus can be converted at sea to become a tethered remotely operated vehicle (ROV) to enable close-up imaging and sampling. The ROV configuration incorporates a lightweight fiber-optic tether (for high-bandwidth, real-time video and data telemetry to the surface), an electro-hydraulic manipulator arm, and sampling instruments. The Nereus vehicle is designed to render all parts of the Earths seafloor accessible to oceanographic science.

Preliminary results in experimental identification of 3-DOF coupled dynamical plant for underwater vehicles

Underwater vehicle control research is presently limited by the availability of explicit experimentally validated plant models. We report an experimentally identified 3 degrees of freedom (DOF) coupled non-linear finite-dimensional plant model for the Johns Hopkins remotely operated vehicle (JHUROV) for the forward, lateral, and heading degrees of freedom. We report the parameter estimation methodologies of total least squares (TLS), ordinary least squares (OLS), and their underdetermined variants. We report that in general TLS is superior to OLS when error exists in all data. We report a 3-DOF coupled non-linear finite-dimensional plant of an underwater vehicle estimated by TLS and compared to the same plant identified by OLS by comparing the mean absolute error between the velocity profile of a numerical simulation and the experimental velocity. We report that the TLS estimate of the parameter model including a fully parameterized quadratic drag performed best overall in cross validation and that in general TLS estimate performs better then OLS estimate. Our goal is to enable the development of new model based controllers capable exact position and velocity trajectory tracking of the forward, lateral and heading DOFs.

Preliminary experiments in fully actuated model based control with six degree-of-freedom coupled dynamical plant models for underwater vehicles

This paper reports a comparative experimental evaluation of one non-model-based proportional derivative (PD) 6 degree-of-freedom (6-DOF) controller and one model-based 6-DOF controller designed to enable fully actuated underwater vehicles to perform 6-DOF set-point regulation and trajectory tracking. We show analytically the non-model-based PD controller is capable of performing locally asymptotically stable set-point regulation. We show analytically the model-based controller is capable of locally asymptotically stable 6-DOF trajectory tracking. Experimental results with the Johns Hopkins University remotely operated vehicle (JHU ROV) show the model-based controllers mean absolute position and velocity tracking error is significantly smaller than the non-model-based PD controller for coupled maneuvers.

Preliminary experiments in comparative experimental identification of six degree-of-freedom coupled dynamic plant models for underwater robot vehicles

This paper addresses the modeling and experimental identification of six degree-of-freedom (6-DOF) coupled nonlinear second order plant models for low-speed, fully-actuated, and neutrally buoyant open-frame underwater vehicles. We report a comparative experimental evaluation of six different candidate plant models whose unknown plant parameters are estimated from data obtained in free-motion vehicle trials. We report an experimental evaluation of the performance of each of the six different 6-DOF coupled non-linear finite-dimensional plant models for underwater vehicles estimated by total least squares (TLS) by comparing the mean absolute error between the experimentally observed vehicle velocities and the velocities obtained by a numerical simulation of the experimentally identified plant models. We also report a cross-validation which evaluates the ability of a plant model to accurately reproduce observed plant velocities for experimental trials differing from the trial from which the plant model parameters were estimated. We conclude that plant models including fully parametrized coupled quadratic drag terms perform best overall in cross-validation. This study has the following contributions: It is the first reported experimental 6-DOF fully-coupled plant model identification and cross-validation of low-speed, fully-actuated, and neutrally buoyant underwater vehicles; it is the first experimental 6-DOF plant model identification for this class of underwater vehicles during free-flight experiments; and it is the first reported use of TLS to perform 6-DOF experimental plant model identification of an underwater vehicle.

Fully actuated model-based control with six-degree-of-freedom coupled dynamical plant models for underwater vehicles

This paper reports a comparative experimental evaluation of one non-model-based proportional derivative (PD) six-degree-of-freedom (6-DOF) controller and two model-based 6-DOF controllers designed to enable low-speed, neutrally buoyant, and fully actuated underwater vehicles to perform 6-DOF set-point regulation and trajectory tracking. We show analytically that the non-model-based PD controller provides locally asymptotically stable set-point regulation, and we show analytically that the model-based controllers provide locally asymptotically stable 6-DOF trajectory tracking. Numerical simulation studies are reported that corroborate the analytical stability results. We report the first comparative experimental evaluation of three different control algorithms for dynamic 6-DOF trajectory tracking of fully actuated underwater vehicles. Experimental results with the Johns Hopkins University remotely operated vehicle (JHU ROV) show that the model-based controllers’ mean absolute position and velocity tracking error is significantly smaller than the non-model-based PD controller for coupled maneuvers. The model-based controllers are shown to outperform the non-model-based controllers over a wide range of variations in the magnitude of derivative feedback gain. The velocity tracking error of the model-based controllers is shown to be on the same order of magnitude as the measurement error of the velocity sensing instrumentation.

Preliminary experiments in nonlinear model-based tracking control of underwater vehicles with three degree-of-freedom fully-coupled dynamical plant models

This paper reports a comparative experimental evaluation of one non-model-based proportional derivative (PD) 3 degree-of-freedom (DOF) controller and two model-based 3-DOF controllers designed to enable a fully actuated underwater vehicle to perform exact trajectory tracking of the X, Y, and heading degrees-of-freedom. We show the non-model-based controller is capable of performing globally asymptotically stable set-point regulation. We show the model-based controllers are capable of asymptotically stable trajectory tracking. The reported controllers were experimentally evaluated on the Johns Hopkins University remotely operated vehicle (JHUROV). We report the model-based controllers mean absolute position and velocity tracking error is significantly smaller than the non-model-based PD controller for coupled maneuvers.

A System for Real-Time Spatio-Temporal 3-D Data Visualization in Underwater Robotic Exploration

This paper reports the development of a real-time human-computer interface (HCI) system that enables a human operator to more effectively utilize the large volume of quantitative data (navigation, scientific, and vehicle status) generated in real-time by the sensor suites of underwater ro botic vehicles. The system provides interactive 3-D graphical interfaces that display, under user control, the quantitative spatial and temporal sensor data presently available to pilots and users only as two-dimensional plots and numerical dis plays. The system can presently display real-time bathymetric renderings of the sea-floor based upon vehicle navigation and sonar sensor data; vehicle trajectory data; a variety of scalar valued sensor data; and geo-referenced targets and waypoints. We report the accuracy of the real-time navigation and bathymetric sonar data processing by comparing the real-time sonar bathymetry of our test tank floor to a high-resolution laser scan of the same tank floor. The real-time sonar bathymetry is shown to compare favorably to the laser scan data

Application of the Biosonar Measurement Tool (BMT) and Instrumented Mine Simulators (IMS) to exploration of dolphin echolocation during free-swimming, bottom-object searches

Data have been collected from two dolphins engaged in free-swimming, bottom-object searches in conjuction with the deployment of both the BMT and IMS systems. Preliminary results demonstrate distinctive search strategies between the animals. The instrumentation described herein has been designed to provide quantitative data for free-swimming dolphins conducting mine-hunting tasks. The data allow researchers never before seen insight into what the dolphin actually does when free swimming and hunting for mine shaped targets. The richness of the data is unparalleled and requires continued in-depth analysis. In addition, when researchers have hypotheses on ways to improve mine hunting sonar, developed algorithms can be applied to the BMT and IMS data to demonstrate how the algorithm performs relative to the dolphin.

Nonlinear Model-Based Tracking Control of Underwater Vehicles With Three Degree-of-Freedom Fully Coupled Dynamical Plant Models: Theory and Experimental Evaluation

This paper reports a comparative experimental evaluation of one model-free proportional derivative (PD) three degree-of-freedom (DOF) controller and two model-based three-DOF controllers designed to enable low-speed, neutrally buoyant, and fully actuated underwater vehicles to perform trajectory tracking in the X, Y, and heading DOFs. We show the model-free PD controller provides locally asymptotically stable set-point regulation. We show the model-based controllers provide locally asymptotically stable trajectory tracking. The reported controllers were experimentally evaluated on the Johns Hopkins University remotely operated vehicle. We report the model-based controller’s mean absolute position and velocity tracking error is significantly smaller than the model-free PD controller for coupled maneuvers. We report the fixed, model-based controllers performed best using a parameter model including fully parameterized quadratic drag model parameters, and that the choice of parameter model has a significant effect on the performance of the model-based controllers during disparate experimental trials.