Patrick McDowell

UUV teams, control from a biological perspective

Remote Operated Vehicles (ROVs) are used extensively for salvage operations, ocean floor surveying and numerous inspection activities that support a wide range of underwater commercial activities. In deep water (greater than 1000 ft) an ROV is the platform of choice because of the depth and endurance limitations for human divers. The key disadvantage to an ROV is the requirement for the long tether. The tether greatly inhibits the speed of the ROV, requires a ship with deck gear capable of handling this cable, and significantly restricts ship movement while deployed. Un-tethered Unmanned Underwater Vehicles (UUVs) have entered the commercial market and have demonstrated the ability to perform deep-water surveys faster and cheaper than towed vessels. With further technological advances, UUVs have the potential for supplementing and even replacing ROVs for many deep-water operations because of the cost and problems associated with the tether. One promising scenario for the near future is to use an ROV or surface ship to control multiple UUVs in a local work area. Typically in this scenario the UUVs are used to extend the sensor footprint of the ROV or surface ship. Another area of interest is the UUV team concept. A stereotypical UUV team would be a heterogeneous mix of several low-cost specific purpose vehicles, guided and supported by one or two higher cost control vessels. Because of the severe restrictions that the sub-sea environment places on communication and positioning, precision underwater navigation is difficult. Currently most precision underwater navigation relies on some sort of infrastructure such as surface ships or underwater beacons placed in known positions. Using these assets as reference-points sub-sea navigation is carried out. Some situations require that the environmental and/or commercial attributes of an area be assessed before an infrastructure exists. In order to do this the UUV team must be able to navigate to an area, carry out its task and return without any pre-existing infrastructure or step by step guidance. Given basic assumptions about the type and frequency of sensor input we present a biologically inspired, decentralized methodology for safely and efficiently moving a loose formation of UUVs to and from the task area with the goal of minimizing outside guidance.

Formation maneuvering using passive acoustic communications

Interest in the use of unmanned underwater vehicles (UUVs) for both commercial and military uses is growing. Control of UUVs poses a difficult problem because traditional methods of communication and navigation, i.e. radio and GPS, are not effective due to the properties of seawater. Control and communication algorithms were developed to carry out multiple UUV formation maneuvering using acoustic communications and first tested in computer simulation and then on mobile robots. Three control schemes, classic logic, behavior, and neural network were tested in line formations in both simulator and lab environments. Results and issues are discussed along with future directions.

Relative positioning for team robot navigation

The research presented in this paper approaches the issue of robot team navigation using relative positioning. With this approach each robot is equipped with sensors that allow it to independently estimate the relative direction of an assigned leader. Acoustic sensor systems are used and were seen to work very effectively in environments where datum relative positioning systems (such as GPS or acoustic transponders) are typically ineffective. While acoustic sensors provide distinct advantages, the variability of the acoustic environment presents significant control challenges. To address this challenge, directional control of the robot was accomplished with a feed forward neural network trained using a genetic algorithm, and a new approach to training using recent memories was successfully implemented. The design of this controller is presented and its performance is compared with more traditional classic logic and behavior controllers.

Kalman filtering with neural networks for change detection in simulated sidescan sonar data

Sidescan sonars produce acoustical imagery which is used to detect bottom objects and characterize features of the seafloor. Change detection is a method that can be used to flag new bottom objects which were not detected during previous sidescan sweeps. Improvements in object detection and classification are critically needed to improve change detection methods. Adaptive filtering techniques may be used to identify objects known or previously marked from historical data, and flag new objects detected as changed. In this study Kalman Filter techniques will be used to estimate weights for a supervised feed‐forward perceptron neural net classifer. Both noise‐free and noisy data are considered. Preliminary results from these techniques using simulations that model sidescan sonar data sets will be presented.

Multi-robot position tracking

Accurate navigation is just one of the many challenges for successfully coordinating multiple robot interaction. It is especially important when trying to quantify the success of new techniques being developed to achieve coordinated formation maneuvering. This paper presents the experimental procedures followed while determining robot navigation error along with an evaluation of the resultant measurements. An inexpensive, easily configurable, camera system is presented that shows the potential to provide accurate position information. Along with a description of the system configuration, test procedures and test dat a are presented and evaluated. Finally, a comparison of the robot navigation error to that of the proposed camera system is presented.

Intervessel navigation using range and range rate

A fundamental requirement for groups of unmanned underwater vehicles to work cooperatively together is the ability for each vessel to know the relative position of its neighbors. While external communication and positioning infrastructures can be used for this purpose, a more flexible approach is to give each vessel the ability to independently discern the location of its neighbors using its own communication and sensor systems. Recent developments in acoustic modems have included the ability to measure range between two vessels as well as the Doppler imparted on the signal by the relative motion between vessels. The measured Doppler shift can be used to compute the instantaneous range rate between two vessels; researchers at NRL are presently working to validate these measurements using GPS. Approaches to vessel relative positioning and navigation using successive range and range‐rate measurements, along with directed vessel maneuvers, will be presented.

A flexible client/server application for robotic control

Flexibility is essential during the research, development and testing of multi-robot systems. This paper describes a flexible client/server application and presents examples of how this flexibility is being used. A software architecture is presented that allows multiple analysis tools the ability to control robotic platforms that have different hardware control systems but use similar locomotion control commands and feedback. Two different robotic platform specifications, one for a wheeled robot and one for a boat, are given along with a description of how both are controlled. An example of how both platforms are being used to conduct research is illustrated in order to show some of the advantages of having flexible robotic configuration and control. Various aspects of the applications evolution while supporting Unmanned Underwater Vessel (UUV) team research are presented along with a discussion of future work to be done.