IEEE Robotics & Automation Magazine - September 2015 - 74

need for interaction with divers arises. Even though human divers today are increasingly being replaced with autonomous underwater vehicles in tedious tasks such as mapping and
searching, there are still many applications that require a human
presence underwater. These applications are mostly related to
unconventional, nonrepetitive tasks such as underwater
interventions.
Underwater ROVs have been commonly used in tandem
with divers, mostly for the purpose of monitoring divers from
the surface. However, the use of autonomous marine vehicles
has taken place only recently. The list of autonomous underwater robots used for diver-robot applications is fairly short:
● AquaRobot [3], developed at McGill University, initially
used as an amphibious robot for exploring underwater en-

GPS
Wi-Fi Link

ASV

Acoustic Link

Diver

Figure 1. A novel concept for increasing diver safety: an
autonomous surface platform, PlaDyPos, serving as the diver's
private satellite and interfacing with the diver through an
underwater tablet.

IEEE ROBOTICS & AUTOMATION MAGAZINE

September 2015

vironments, was used to track divers based on visual detection of their motion [4].
● The BUDDY AUV, developed for the purposes of the
European project Cognitive Autonomous Diving Buddy
(CADDY, http://caddy-fp7.eu/), is the first AUV for interacting with divers by using an underwater touchscreen.
Tracking and navigating divers using ASVs was first addressed by the European project Cooperative Cognitive
Control for Autonomous Underwater Vehicles [5]. A fleet of
three ASVs was deployed, and successful diver-tracking and
-guidance experiments were performed by using single range
measurement from the vehicles [6]. This project was a stepping
stone toward the CADDY project, under which the results presented in this article were obtained. It should be mentioned that
other efforts have been made toward robot-diver interaction,
such as [7]; however, these were done only in the simulation environment, without addressing the issues of diver detection or
localization. While using acoustic positioning systems to localize divers is the most straightforward method, visual [4], [8]
and sonar [9] detection can be found in the literature.
Contributions
The main contributions of this article are both in technical and
control aspects as well as benchmarking and experiment design and execution. The first contribution is the development
of the diver-tracking system consisting of an autonomous surface marine vehicle and an underwater diver interface used for
two-way communication between the diver and the surface
vehicle. The complete system is described in the next section.
The second contribution is the onboard diver-tracking algorithm that uses intermittent acoustic diver position measurements fused with a diver motion estimator. The models,
control, and tracking algorithms were initially described in a
previous paper by the authors [10], but they are also included
in this article for completeness. The third contribution is the
design of a benchmark scenario with associated metrics for
human-robot tracking performance measurement in the underwater environment. The benchmark scenario enables replicability of experiments in real conditions with human divers.
Finally, the fourth contribution is the design and execution of
an experimental plan that allows the identification of uncertainties introduced by the human diver, environment, unmodeled dynamics, and acoustic sensors by using tracking
performance metrics.
Diver-Robot System Description
The overall diver-tracking system presented in this article
consists of three main components: 1) the diver-tracking marine ASV, 2) the diver, and 3) an underwater tablet carried by
the diver, serving as the diver's interface. The components are
described in the following section, while the communication
scheme between them is shown in Figure 2.
ASV-PlaDyPos
The ASV PlaDyPos (named after its initial purpose as a platform for dynamic positioning), Platform for Dynamic

http://www.caddy-fp7.eu/

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - September 2015