The SWIMMER: a System for underWater Imaging and Monitoring for Marine
Environment Research
Abstract
The Basking Shark (Cetorhinus maximus) is an elusive creature
that possesses a highly-specialized filter-feeding system capable of
filtering around two million liters of water every hour. Learning from
such filters could be instrumental in making new filtration systems for
microplastics or pollutants in the future. However, the specific
filtration mechanism, as well as many other aspects of their morphology
and behavior, is not well understood. There are numerous difficulties
involved with the study of basking sharks, particularly finding the
sharks in their ocean habitat and observing them without disturbing
them. Here, we present a new open-source robotic platform, the System
for underWater Imaging and Monitoring for Marine Environment Research
(SWIMMER, SWIM for short), for non-invasive underwater imaging. The
robot was designed to produce high-resolution stereo images at both
surface level and underwater. The imaging payload is carried by a
surface robot to make it easy to monitor, control, and retrieve. The
robot is built low-cost and entirely from off-the-shelf components and
3D printing to enable us to scale up the robot to a swarm in the future
for imaging from multiple points of view and for other marine
applications. To demonstrate the platform, we show its use in recent
field trials where we were able to collect video data of 9 different
individuals in 6 different encounters, 3 encounters involved 2 sharks,
as well as collecting data about its capabilities. This work has the
potential to provide new insights into basking shark feeding behavior,
as well as potentially making marine swarm robotics more accessible to
researchers everywhere due to the low cost and ease of construction of
the SWIMMERs.