Marine scientists and engineers recently set free a robot called Sirius to explore the ocean floor at Ningaloo in WA. This Autonomous Underwater Vehicle (AUV), equipped with high resolution cameras and sonar and the ability to navigate on its own, has allowed them to gain unparalleled insights into deep water environments. The joint expedition between scientists from the Australian Institute of Marine Science (AIMS) and the University of Sydney's Australian Centre for Field Robotics (ACFR) was designed to explore the suitability of AUVs for environmental monitoring.
"This has been a major step forward in understanding the deeper water seabed communities that occur around Ningaloo. In the past, ecologists have relied on trawl gear and tools that are fixed to the support vessel by cables to collect information about the seascape at great depths" says expedition leader, Max Rees from AIMS. "The fact that the vehicle does not require a tether gives us a great deal of flexibility in the areas we can work".
BHP Billiton is providing AIMS with research funds to enhance marine knowledge of the North West Cape and Ningaloo Reef and it is this corporate support for science that is largely responsible for bringing the AUV to Ningaloo. The strong corporate patronage is enabling projects based on cutting edge marine technologies to compliment and extend the impact of the Governments’ marine science initiatives at Ningaloo under WAMSI.
Sirius was designed for detailed surveys of underwater environments in depths of up to 700 metres. It uses an onboard computer system and an array of instruments to navigate just metres above the seabed. The vehicle scans the seafloor creating bathymetric sonar maps while collecting thousands of high-resolution digital images per hour. It is programmed prior to deployment to follow a survey pattern while its progress can be monitored on the surface via an underwater acoustic modem, in this case from the AIMS Research Vessel Cape Ferguson.
Dr Stefan Williams from ACFR says, "Last year engineering trials were conducted at Ningaloo and on the Great Barrier Reef but this is the vehicle’s first major scientific survey expedition."
Speaking from the ship in the last days of the expedition Max Rees said, "An unexpected discovery was the abundance of life clinging to the very tops of the steep deepwater canyons, where the continental shelf drops away not far from Ningaloo’s shallow reef. Many of the sponges and deepwater corals that were photographed may be new to science. Also of interest were fields of mound features that were observed on the shallower seabed plateau. We suspected the mounds were there, although their cause remains a mystery, but have now been able to see them in spectacular detail."
In all, half a million images were collected on 20 dives, with the vehicle operating within canyons in depths of up to 250m. Captured directly onto onboard computers, the individual stereo images can be used to measure minute features of the seabed with unmatched clarity and detail. The researchers are assembling these images into mosaics that enable them to observe larger scale patterns in the data. "Stitched together the digital photos will effectively provide photographic maps of the seabed, giving an idea of the distribution of benthos as if the overlying deep ocean had been removed," Dr Williams says.
"For AIMS, this expedition has played a critical role in exploring the use of robots in underwater research. Robotic technology is a very effective way to run controlled surveys of steep walled canyons and other complex terrain. The vehicle uses a variety of sensors to figure out where it is and to follow its planned mission. The data collected can then be geo-referenced to allow scientists to assess how deep sea communities are distributed as a function of depth or seabed composition. For industry this technology may be used to ensure the environment is adequately surveyed before projects such as underwater pipelines are commissioned. Worldwide, companies that lay deep sea cables and conduct geophysical and hydrographic surveys are increasingly relying on AUVs," Dr Williams says.
"The next step is to look at repeating surveys to assess changes in these habitats over time. Given the current deployments, we can send the vehicle down to have another look in six months or a year. To observe really fine scale change will require that the robots have the ability to recognise parts of our survey when we return." Parallel research into data management and artificial intelligence are part of the R & D program. "We are working on automated methods to deal efficiently with the massive volume of data we are collecting. This is important to allow us to quickly show the scientists the most interesting images or seafloor structures in the areas we have surveyed. It also raises the possibility of allowing the vehicle to decide during its mission when it is seeing something of particular interest. It might then be instructed to go in and have a closer look when it sees something out of the ordinary." explains Dr Williams
Editor's Note: Original news release can be found here.