P012

Command, Control and Data Transport for Underwater Robots Using

Acoustic Communications

A0023 Dale Green Benthos, Inc.

A0024 Chuck Bernstein Naval Coastal System Station

Undersea robots are being used in environments where immediate

supervision by human operators becomes ever more difficult and

potentially dangerous. This necessitates the evolution toward

autonomy in the robot, while at the same time maintaining some

capability to provide a reduced level of supervision, and transmission

of sensor data to the human observer. The US Navy's Coastal Systems

Station (CSS) is developing battery-powered robotic crawlers for mine

countermeasures purposes in very shallow water. These machines carry

a variety of sensors designed to detect the presence of underwater

objects and to classify them as either man-made or natural, as well as

camera systems to capture images of the objects. Benthos, Inc. is

collaborating with CSS to provide acoustic communications (acomms) for

both command and control and for transmission of images and sensor

data. The University of California at San Diego is developing

Wavelet-based image compression to support the telemetry.

Experimental results demonstrate that compression ratios of better

than 100:1 provide acceptable visual representation to a human

operator. The modem aboard the crawler provides robust acomms at

rates up to 1200 bits per second at kilometer ranges. Typical 320x240

pixel greyscale images, compressed at 100:1, can be transmitted in

under 6 seconds. In the near term, the modems will provide the same

image in under 2 seconds. Using existing capabilities, the acoustic

transmission can be relayed via multiple modems for many kilometers,

then via a Benthos gateway buoy, to the operator via a spread spectrum

radio or cell phone link. Current efforts at CSS focus on the

evolution of crawler autonomy with occasional supervision supplied via

the acomms link. Sensors are being integrated with the vehicles to

enable them to detect, locate, and image objects such as mines on the

sea floor. The crawling robots have the distinct advantage of

operating in the same plane as the targets, and can therefore exploit

target features that cannot be used in other countermine approaches.

This work is funded by the Office of Naval Research, Code OE 321.