Home | Contact ST  

Feature Article

Autonomy for Hull Inspections, Underwater Mine Neutralization
SeeByte Autonomy Software Reduces Operator Workload and Improves Situational Awareness of Unmanned Underwater Vehicles

AUTHORS:
Scott Reed
Head of Engineering
Jon Wood
Vice President of U.S. Operations
Dr. Ioseba Tena
Sales Manager
SeeByte Ltd.
Edinburgh, Scotland

Underwater mine countermeasure (MCM) operations are now routinely using UUVs. Commercial off-the-shelf products are available that are man-portable and can carry multiple payloads, including side scan sonar, forward-looking sonar, video and synthetic aperture sonar sensors. UUVs can conduct their operations close to the seafloor, are easily deployable and provide high-resolution sensor data, enabling the detection and classification of mine-like threats. Navigation sensors now provide solutions that allow detected threats to be localized and reacquired. These systems offer an MCM solution that removes humans from the battlespace and complements traditional ship-mounted MCM technology.

While UUVs are autonomous in that they operate in-mission without an operator in the loop, they are still generally used as data collection devices. A UUV’s mission is programmed before the mission as a series of waypoints, after which the UUV data are processed and analyzed post-mission by a human operator. Any UUV-based response to anything found in the sensor data requires the UUV to be redeployed, and the scope of the mission is again limited to the UUV carrying out predetermined waypoints.

SeeByte is at the center of efforts to increase the autonomy capabilities of UUVs. Autonomy is based on the vehicle being able to continually perceive its environment and dynamically and appropriately react while still ensuring that the high-level goals of the mission are fulfilled. UUV mission planning is transitioning from operator-driven waypoint planning, where the vehicle is told exactly where to go, to goal-based mission planning. In this scenario, the responsibility for executing, monitoring and adapting the mission sits with the onboard autonomy system. This shift in responsibility also allows the system to be extended to include multiple collaborative vehicles.

SeeByte is actively involved with the U.S. Office of Naval Research (ONR) in several programs aiming to demonstrate autonomy capabilities within an MCM context. The Confined Area Search program is focused on producing an autono­mous hull inspection capability using the Bluefin Robotics Corp.’s (Quincy, Massachusetts) Hovering AUV (HAUV). The littoral neutralization system program aims to incorporate autonomy capabilities onto the U.S. Navy Common Neutralizer vehicle.


Autonomy and Automatic Target Recognition
Onboard autonomy is dependent on the vehicle being able to accurately sense its environment. If the vehicle cannot monitor its surroundings, it cannot adapt to change and would be restricted to following preprogrammed waypoints. Onboard automatic target recognition (ATR) modules allow the UUV to detect, track and monitor features of interest, dynamically controlling the mission.

In the military context, the feature may be a mine that the UUV has been asked to inspect. It could also be a feature on a ship hull, such as a propeller shaft, a harbor wall or a diver that needs to be followed. Once the ATR is robustly able to detect the target, the onboard autonomy modules may use this information to control and direct the vehicle relative to the object of interest. The vehicle may autonomously inspect the object, factoring in external factors such as sea currents based on feedback from the ATR.

Within the underwater domain, sonar is the most common sensor used onboard UUV systems. Forward-looking sonar technology is typically used within both the hull inspection and mine neutralization domain to locate possible mine threats. SeeByte’s ATR system has been developed to be sensor-independent, and it has been used in both of these domains to detect, classify and localize mine-like targets. The algorithms must be capable of operating in real time so that possible targets can be detected in-mission. The onboard autonomy modules then can react immediately to this information.

When water visibility allows, video ATR technology may be used to complement the information from the sonar and further verify the identity of a given target. It may also be used to aid near-range control behaviors when the object of interest sits within the blanking distance of the sonar. The SeeByte ATR system used a supervised learning approach based on novel pattern recognition and computer vision techniques, which allow the model to process the video images in real time.


Mine Neutralization
SeeByte has been actively involved in ONR mine neutralization for many years, initially demonstrating in June the onboard autonomy capabilities on board iRobot’s (Bedford, Massachu­setts) Transphibian vehicle. These capabilities are under consideration for transition onto the U.S. Navy Common Neutralizer Archerfish Vehicle, developed by BAE Systems (London, England) and integrated into the U.S. Navy Airborne Mine Neutralization System by Raytheon Co. (Waltham, Massachusetts).

Once the vehicle is brought into sonar range of the target, the autonomy modules initiate a search and reacquire behavior to locate the target. During the reacquire pattern, the sonar ATR module actively searches for the target. Once the target is initially detected, confidence in the detection is built up over time by comparing the automated detection to the believed location of the mine threat. This location information is typically obtained during the earlier search-classify-map phase and the reacquire and identify phases of the MCM operation. Upon confirmation that the correct target has been localized by the ATR, the autonomy modules home the vehicle toward the target and initiate a target-relative hover. This capability was demonstrated throughout multiple trials in 2011 against both volume and bottom targets in realistic environments.

The onboard autonomy modules can carry out object-relative maneuvers as required. The complexity of these maneuvers is limited by the control capabilities of the specific vehicle, which is typically determined by the vehicle’s specific thruster configuration. The object-relative control is dependent on the sonar and video ATR systems onboard the vehicle, providing a robust, repeatable estimate on the location of the target. To continue this article please click here.



Dr. Scott Reed earned a master’s degree in astrophysics after which he completed his Ph.D. at the Ocean Systems Laboratory at Heriot-Watt University. He joined SeeByte in 2004. In 2005, he was an invited scientist at the NATO Undersea Research Centre. In 2006, he oversaw SeeByte’s involvement in U.S. Office of Naval Research programs. In March 2009, Reed became SeeByte’s head of engineering.

Jon Wood joined SeeByte Ltd. in August 2007 as vice president of its U.S.-based operations and established SeeByte Inc. in 2008. His naval and joint operations background and extensive international and interagency experience provide significant insight and expertise into both U.S. and international defense and national security leadership, decision making and procurement processes.

Dr. Ioseba Tena is responsible for the development of SeeByte’s commercial strategies and managing the marketing sales process. He has been involved in developing smart solutions for the underwater vehicle industry for more than 10 years and continues to provide engineering expertise to the team.




-back to top-

-back to to Features Index-

Sea Technology is read worldwide in more than 110 countries by management, engineers, scientists and technical personnel working in industry, government and educational research institutions. Readers are involved with oceanographic research, fisheries management, offshore oil and gas exploration and production, undersea defense including antisubmarine warfare, ocean mining and commercial diving.