279
perception of the vehicle. Sensor technology is advancing rapidly, driven by miniaturization, fiber optics, and
low-power electronics. Sensors span a broad domain of
transduction techniques and signal processing, and can
be categorized in various ways. Non-acoustic sensors,
in particular, have progressed significantly in recent
years, and Chap. 18 reviews this field of development.
There are many advantages of using multiple autonomous ocean vehicles to collaboratively accomplish
a mission. Four chapters (1922) examine various aspects of this approach. Chapter 19 reviews cooperative
control of autonomous ocean vehicles for environmental monitoring. An array of vehicles can act as a mobile,
reconfigurable, large aperture sensor (antenna). Using
feedback control to coordinate relative vehicle motion,
such an array can adapt efficiently to a changing environment. The methodology and advantages of such
cooperative control are described in two unprecedented
field experiments in the coastal ocean. Chapter 20
describes the concept of nested autonomy as an architecture for distributed ocean sensing. Hierarchical,
distributed command and control enables deployment
of an array of autonomous ocean vehicles over a wide
area for a long period of time with little or no human supervision. Effective coverage, limited communication
range, safe operation, and environmental uncertainty
are all primary constraints. This approach provides
a mechanism for balancing autonomy with the periodic
need for operator intervention. Chapter 21 focuses on
time-optimal path planning and adaptive sampling for
groups of ocean vehicles in realistic ocean conditions.
Path planning optimization and adaptive sampling both
involve prediction of ocean variables, and thus ocean
modeling is essential. The state-of-the-art using dynamical models in feedback control of autonomous ocean
vehicles is documented for both idealized and realistic environments. The goal is to improve the predictive
skill of the continuous (space and time) ocean fields
being measured. Optimal path planning and adaptive
sampling can also be used to improve the predictive
skill of detecting and tracking discrete objects (targets)
in the environment. In many sport, military, and biological endeavors, cooperative strategies have proven
themselves to be advantageous over non-cooperative
ones. For a multi-vehicle team operating in the continuous and transient ocean, cooperative behavior involves
optimizing a high-dimensional parameter space, even
more so in cases when the target has the ability to make
intelligent choices to avoid being tracked. Chapter 22
examines recent progress to construct a theoretical
framework and initial applications for maritime surveillance. A methodology is developed that can help bridge
the gap between a top-down view starting with theoretical concepts and a bottom-up view based on the details
of real experimentation and execution at sea.
Deploying autonomous vehicles for extended periods in the coastal ocean requires safe and reliable
operation in potentially congested waterways. The existing rules of the road for marine vessel operation are
evolving to accommodate autonomous vehicles. Simultaneously, designers of such vehicles are developing
advanced autonomy behaviors that exhibit human-like
performance. Chapter 23 provides a snapshot of this
rapidly changing operational environment, with particular attention paid to the software shaping high-level
autonomy and on the legal framework for acceptable integration into public water space. In contrast to coastal
ocean operation, where collision is a principal risk, the
risk in deploying autonomous vehicles for extended
periods in the open ocean lies more in their potential disappearance. Indeed, testing of true autonomy
has often been limited by risk-of-loss aversion. When
evaluating risks, various stakeholders may weigh consequences differently. For example, the owner may
focus on the risk of vehicle loss, whereas the user on
risk of data loss. Other risks, such as hazards to navigation, affect all stakeholders. Chapter 24 describes
a risk management process using several methods tailored specifically to deployment of autonomous ocean
vehicles.
perception of the vehicle. Sensor technology is advancing rapidly, driven by miniaturization, fiber optics, and
low-power electronics. Sensors span a broad domain of
transduction techniques and signal processing, and can
be categorized in various ways. Non-acoustic sensors,
in particular, have progressed significantly in recent
years, and Chap. 18 reviews this field of development.
There are many advantages of using multiple autonomous ocean vehicles to collaboratively accomplish
a mission. Four chapters (1922) examine various aspects of this approach. Chapter 19 reviews cooperative
control of autonomous ocean vehicles for environmental monitoring. An array of vehicles can act as a mobile,
reconfigurable, large aperture sensor (antenna). Using
feedback control to coordinate relative vehicle motion,
such an array can adapt efficiently to a changing environment. The methodology and advantages of such
cooperative control are described in two unprecedented
field experiments in the coastal ocean. Chapter 20
describes the concept of nested autonomy as an architecture for distributed ocean sensing. Hierarchical,
distributed command and control enables deployment
of an array of autonomous ocean vehicles over a wide
area for a long period of time with little or no human supervision. Effective coverage, limited communication
range, safe operation, and environmental uncertainty
are all primary constraints. This approach provides
a mechanism for balancing autonomy with the periodic
need for operator intervention. Chapter 21 focuses on
time-optimal path planning and adaptive sampling for
groups of ocean vehicles in realistic ocean conditions.
Path planning optimization and adaptive sampling both
involve prediction of ocean variables, and thus ocean
modeling is essential. The state-of-the-art using dynamical models in feedback control of autonomous ocean
vehicles is documented for both idealized and realistic environments. The goal is to improve the predictive
skill of the continuous (space and time) ocean fields
being measured. Optimal path planning and adaptive
sampling can also be used to improve the predictive
skill of detecting and tracking discrete objects (targets)
in the environment. In many sport, military, and biological endeavors, cooperative strategies have proven
themselves to be advantageous over non-cooperative
ones. For a multi-vehicle team operating in the continuous and transient ocean, cooperative behavior involves
optimizing a high-dimensional parameter space, even
more so in cases when the target has the ability to make
intelligent choices to avoid being tracked. Chapter 22
examines recent progress to construct a theoretical
framework and initial applications for maritime surveillance. A methodology is developed that can help bridge
the gap between a top-down view starting with theoretical concepts and a bottom-up view based on the details
of real experimentation and execution at sea.
Deploying autonomous vehicles for extended periods in the coastal ocean requires safe and reliable
operation in potentially congested waterways. The existing rules of the road for marine vessel operation are
evolving to accommodate autonomous vehicles. Simultaneously, designers of such vehicles are developing
advanced autonomy behaviors that exhibit human-like
performance. Chapter 23 provides a snapshot of this
rapidly changing operational environment, with particular attention paid to the software shaping high-level
autonomy and on the legal framework for acceptable integration into public water space. In contrast to coastal
ocean operation, where collision is a principal risk, the
risk in deploying autonomous vehicles for extended
periods in the open ocean lies more in their potential disappearance. Indeed, testing of true autonomy
has often been limited by risk-of-loss aversion. When
evaluating risks, various stakeholders may weigh consequences differently. For example, the owner may
focus on the risk of vehicle loss, whereas the user on
risk of data loss. Other risks, such as hazards to navigation, affect all stakeholders. Chapter 24 describes
a risk management process using several methods tailored specifically to deployment of autonomous ocean
vehicles.