IEEE Robotics & Automation Magazine - June 2019 - 95

Current Capabilities and Challenges
Over the years, the capabilities of UUVs have steadily
increased, from ROVs to AUVs and now I-AUVs and
H-ROVs. Transitioning from ROVs to AUVs pushed new
developments in localization, autonomy, and communications.
H-ROVs and I-AUVs helped drive new manipulation capabilities. Software development has been made easier by middleware frameworks such as Mission Oriented Operating
Suite-Interval Programming or, more recently, the Robot Operating System. Simulators, such as Gazebo or the UnderWater
Simulator, obviate the need for the complexity and cost of at-sea

testing. However, while progress has been steady, some important challenges remain.
In the following sections, we first explain the current state of
the art by grouping the most successful and relevant approaches
for each capability and then present the remaining challenges.
We conclude with a short summary exemplifying the evolution
of robots and capabilities through the years.

WIKIPEDIA

Sensing
The range of sensors that AMVs and UUVs can carry as payload is very wide and depends on the application. We focus
here exclusively on sensors used by the robot to extract the
information needed to perform its mission safely and efficiently. The most common sensors used in UUVs are optical
and acoustic sensors. Optical cameras can be used for many
applications, such as object detection and classification, 3D
reconstruction, mosaicking, and simultaneous localization
and mapping (SLAM). They can also be used for obstacle
avoidance or keeping a constant distance from the bottom
for safety reasons. Similarly, for inspection tasks, optical
cameras can help UUVs survey the area without coming into

Figure 1. A typical work-class ROV used in industry today.

BLUEFIN

inspection, repair, and maintenance. They range from small,
portable vehicles to large work-class systems. Most of today's
oil and gas operations and marine science would be impossible
without them.
The 1990s saw the development of torpedo-shaped AUVs
to provide fast, large-scale, high-resolution surveys of the seabed [1] (see Figure 2). This new technology removed the need
for a tether (required by ROVs) and their associated support
ship. It also provided higher-resolution data than traditional
approaches as well as access to restricted areas, for instance,
under ice or in shallow waters. These vehicles found widespread applications in defense [2], oil and gas, and cable surveying [3]. However, they were not able to interact with
structures for close-up inspection or manipulation. Moreover,
while AUVs have commercially been used routinely for surveying [4], surveillance [5], and, recently, inspection [6], their
economic benefit is greatly reduced if another system (typically an ROV) and an expensive ship are required to perform the
intervention resulting from the inspection. Therefore, the next
efforts in UUV technology focused on the development of
intervention AUVs (I-AUVs) and hybrid ROV-AUVs
(H-ROVs) endowed with manipulation capabilities. This is
still an area of ongoing research and probably the key technology development required to move AUVs into the mainstream
of subsea operations.
The United States led the way early on in the mid-1990s
with one-degree-of-freedom (1-DoF) I-AUVs [7]. In the
late 1990s, the Advanced Manipulation for Deep Underwater Sampling European Union (EU) project developed
the first robust 7-DoF electrical robot arm [8]. During the
2000s, the first autonomous intervention [9] and first freefloating intervention within the milestone Semi-Autonomous Underwater Vehicle for Intervention Mission project
[10] took place. In the last decade, lighter I-AUVs have
been developed, such as the Girona 500 used in several EU
projects (TRIDENT [49], PANDORA [50]). Meanwhile,
free-floating manipulation flourished. For instance, in
TRIDENT, for the first time, a vehicle and manipulator of
similar weight performed autonomous grasping, coordinating the base and arm movements [11]. Dual-arm
manipulation was achieved initially in the early 2000s [12],
but, currently, there are research projects dealing with
dual-arm, free-floating underwater manipulation using
H-ROVs [13], [14].

Figure 2. A typical survey AUV.

JUNE 2019

IEEE ROBOTICS & AUTOMATION MAGAZINE

IEEE Robotics & Automation Magazine - June 2019

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - June 2019