Spring 2021 Marine Autonomy Roundtable Schedule
- All events start at 930 EDT, with webinar open at 925 EDT
Participants:
Dr. Gabriele Ferri
NATO Center for Maritime Research and Experimentaton
Marine robotic networks for surveillance: the importance of autonomy and cooperation in communications-limited environments
Recent developments in marine robotics have made small and low-cost Autonomous Underwater Vehicles (AUVs) and, more in general, Maritime Unmanned Systems (MUS), an opportunity. MUS start to be mature enough to be used at sea for complementing or even substituting traditional means (mainly ship-based operations and/or fixed sensors) in ocean exploration/monitoring.
Compared to traditional assets, these small, low-power and mobile units have usually limited processing and communication capabilities. Nonetheless, when deployed in a spatially separated network, they can be interconnected to form an intelligent network characterised by features of scalability, robustness, reliability and adaptability.
Oceans, however, are an extremely challenging environment for robots operation. Open scientific and engineering challenges remain to be solved, for instance deep-water navigation (GPS cannot be used underwater) or the limited underwater endurance (no combustion engine can be used underwater). The limitations of underwater acoustic communications are a defining feature of the underwater domain. A direct link between a control centre and the robots cannot be guaranteed and inter-robot communications are intermittent and not reliable. In these conditions, autonomy, defined as the capability to make decisions based on the collected data, the environment and the information received from other collaborators, becomes crucial. Autonomy can be pivotal for the deployment of networks of robots.
The talk describes the work conducted at CMRE on these topics. In particular, an Anti-Submarine Warfare (ASW) scenario will be analysed. We will describe different types of hybrid robotic networks under development for surveillance applications. We will present how single-robot and cooperative autonomy strategies have the potential to increase the network performance, supporting data fusion and optimising the node actions.
In a limited communication environment, autonomy is the way to enable the robots to exploit their mobility for adapting their mission to the changing environmental conditions and to the evolving scenario. We conclude reporting the ongoing approaches to develop more environmentally aware robot decision-making.
Dr. Gerben Peeters, Prof. Peter Slaets
KU Leuven
Towards automated inland cargo shipping in Western-Europe
The current Western-European inland waterway transport sector seems to be evolving into an economically unviable transport mode, whereas this sector presently offers a more sustainable solution for hinterland freight transport than road-based transport. Automated or unmanned inland cargo vessels might induce a paradigmatic change in this tension field. Although the conceptual idea of automated or unmanned vessels itself might be straightforward, it is not clear how these vessels could or should look like.
Therefore, the Intelligent Mobile Platforms research group from the KU Leuven explores the technological feasibility of unmanned inland cargo vessels. In this talk, we will give an overview of our past, ongoing, and future projects for automated inland vessels.
Walter Holemans
Marine Robotics, LLC
Autonomous sailing robot for marine observation and ecosystem restoration
Marine Robotics, LLC makes an autonomous sailing robotic catamaran meant as a general development platform for observation of water quality (pH, Salinity, turbidity), surface atmosphere (Temperature wind speed, direction, RH), and marine life (plant and fish counting and health observation). The boat is 8.5 feet long, 7 feet tall and weighs 130 pounds. It travels at about 3 miles per hour. The design is meant for mass production so swarms of hundreds of boats can dynamically observe larger scales, like the salinity variation in tidal areas due to rain events. To advance this technology several unique problems need to be solved. These include image recognition of marine life such as oysters so they may be counted. Another is collision avoidance using inexpensive, low power means which may include, AIS and cameras. Sailing algorithms will admit improvements. One challenge is zig-zagging into the wind while simultaneously avoiding objects like a reef in pursuit of a changing condition like an oil spill for example. Another includes acting as a communications or power node for vertical aircraft (the boat acts as a quadcopter carrier). Similarly, another is to act as a submersible support. Another is the control of winched instruments to conduct vertical profiles in deep water. Another is the control of a fish-tail mimicking propulsion systems. Such an appendage is exciting as it could also be employed in shallow muddy water at the waters edge.
Marine Robotics can contribute functioning boats that would then be employed to test-verify novel algorithms. The solution to these challenges will enable a much more cost effective marine observation and hopefully aid in ecosystem restoration in US coastal waters, rivers, lakes and bays. Environmentalists, aquaculturists, researchers, marine energy, shipping and the government have a growing need for low cost marine observation.
Andrea Munafo, Dr. Scott Reed
SeeByte
The evolution of marine robotics at SeeByte: moving from single sortie missions to multi-vehicle, multi-day campaigns
The achieved maturity of unmanned underwater and surface vehicles (UUVs and USVs - UxVs) makes the deployment of multi-vehicle networks practical and cost-effective. For example, many scientific, civilian and military applications nowadays require the usage of autonomous underwater vehicles (AUVs) and networks of AUVs have been deployed in scenarios ranging from surveillance to mine-counter measurement operations or oceanographic systems.
This talk will describe the evolution of marine robotics at SeeByte. From the early days where a single AUV was the tool of choice and missions were normally planned as a sequence of pre-defined waypoints, to today's complex operations that involve deploying multi-domain robots (air, surface, underwater) over multiple days. Experimental results and lessons learnt will be discussed from recent operational deployments where SeeByte supported operators to efficiently schedule and execute multi-vehicle, multi-days and large area campaigns.
Micheli Michele
NATO Center for Maritime Research and Experimentaton
Introducing the backseat driver paradigm into the Teledyne Slocum Glider – a successful MOOS-ification
The problem of monitoring marine traffic by using passive acoustics is a topic of great interest for the NATO-STO Centre for Maritime Research. The use of a network of autonomous vehicles equipped with innovative acoustic sensors provides a large set of advantages (collaboration, flexibility in configuration and geometry) if compared with more traditional approaches (buoys, moored systems). On the other hand, passive underwater monitoring has demanding requirements (such as covertness, long endurance and very low noise platforms), which are not easily achievable by a traditional propeller driven AUV. Underwater gliders are the ideal platforms to perform this kind of task. CMRE has a long experience with the Teledyne Slocum underwater glider, which has successfully used in the last 15 years to perform oceanographic and environmental acoustic missions.
The Slocum glider has been originally designed for long range missions with a particular focus on power consumption, persistency, cost reduction and reliability. A traditional Slocum can be commanded to perform waypoint or fixed heading missions. This low-level autonomous capabilities are not sufficient for its effective use in passive monitoring missions. This type of missions requires an AUV with the capability to make complex autonomous decisions triggered by the results of signal processing executed on board. The optimal solution to increase the level of autonomy of the Slocum has been identified in the backseat driver architecture adopted by MOOS-IvP, which has already been successfully used on a variety of autonomous vehicles at CMRE. In order to limit the changes to the vehicle, the backseat has been installed on a new payload which, from the glider perspectives, acts as a smart sensor.
The talk describes how the Moos-ification of the Slocum gliders has been successfully achieved at CMRE.
Mikayla Cohen
Sea Machines, Boston MA
Cross-Pollination of Marine and Space Autonomy
This talk will describe how the domains of marine autonomy and space exploration autonomy can benefit from the similarities and technological advancements of one another.
Dr David Battle
Mission Systems, Sydney Australia
Maritime Autonomy Simulation Tricks and Techniques
Mission Systems is developing a variety of physics-based maritime autonomy simulation capabilities, including:
- Acoustic communications simulation
- Sidescan sonar simulation
- ATR synthetic training
- Deep ocean acoustics
The core elements of our simulations include physics engines to compute realistic platform dynamics and acoustic propagation models based on GPU-accelerated ray-tracing. Due to the limitations of game engine technology, one paradigm we settled on early was the decoupling of physical dynamics calculations from sensor modelling. This allows for the distribution of computational load across hardware best suited to each task. On the down-side, this also requires the creation and synchronization of duplicate 3-D environments, including their dynamic aspects.
This talk will cover some of the core methods we use for distributed maritime autonomy simulations, which use MOOS as their underlying communications framework and pHelmIvP as their autonomy engine.
Scott R. Sideleau, Chris W. Gagner, Dr. Michael L. Incze
Naval Undersea Warfare Center, Newport RI, USA
Intelligence Preparation of the Environment (IPOE) Operations Research enabled by Open Systems Architecture
Maritime robotics platforms equipped with Mission Oriented Operating Suite (MOOS) and Interval Programming (IvP) Helm have been employed successfully by academia and public sector research institutions to achieve various applied research goals over the last 15-years. Often, reactive behaviors that respond to through-the-sensor data have been exercised and refined. This work discusses the open systems architecture and standards that have been leveraged, developed, and deployed by NUWC Newport to support various operational demonstrations with the US Navy, US Marine Corps, and its NATO partners. In this case study, we explore the concept of operations (CONOPS) behind Intelligence Preparation of the Environment (IPOE); the evolution of common control segments enabling minimally trained operators, the benefits of equipping unmanned vehicles (UxV) with common communications, discuss one of the developed behaviors enabling IPOE operations research, and opportunities for future collaboration.
Dr. Cameron Matthews
Naval Surface Warfare Center, Panama City FL, USA
Innovative expendable UUV designs and autonomy implications
Dr. James McMahon
United States Naval Research Laboratory, Washington DC, USA
United States Naval Research Laboratory Maritime Autonomy: Single and Multi-Vehicle Control
Dr. Sekhar Tangalira
Lockheed Martin, Florida, USA
Overview of Lockheed Marine Autonomy Programs
Chi-Fang Chen (1) Hsueh-Cheng "Nick" Wang (2), Ching-Tang Hung (1), You-Cheng Zhang (1),
(1) Department of Engineering Science and Ocean Engineering National Taiwan University
(2) Department of Electrical and Computer Engineering National Yang Ming Chiao Tong University
Development of Underwater Acoustic Detection, Classification, and Localization Techniques on marine mammal utilizing Autonomous Surface Vehicle
This research is a three-year project and its goal is development of key technologies for passive acoustic monitoring (PAM) network which includes underwater acoustic detection, localization and tracking of marine mammal utilizing an unmanned surface vehicle. In recent years, our research team is committed to investigate the critically endangered species, the Chinese white dolphin (Sousa chinensis taiwanensis) inhabits the coastal waters of western Taiwan. The development of offshore wind farms overlaps with the habitat of the Chinese White Dolphin, posing a threat to the survival of the Chinese White Dolphin. Therefore, it is necessary and urgent to investigate the population distribution of the Chinese White Dolphin. Thus we introduce PAM network to monitor their population. Key technologies of PAM network include utilizing various platforms with acoustic payloads to conduct underwater acoustic detection, localization, and tracking. Platforms can be unmanned surface vehicle and fixed such as buoy or bottom mounted units. Field test would be conducted in the coastal regions between the river mouth of Dajia River and the north jetty of Taichung harbor. The water depths are between 5 to 20 m where sighting rate of Chinese white dolphins is high. We will first focus on underwater detection of each unit, then the target localization with multi-sensor detection, lastly a field test of target localization and tracking of underwater acoustic detections with multiple platforms are planned.