@article{article, author = {Aya Saad | Department of Engineering Cybernetics, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and Annette Stahl | Department of Engineering Cybernetics, and Centre for Autonomous Marine Operations and Systems, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and Andreas Våge | Department of Engineering Cybernetics, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and Emlyn Davies | Department of Environment and New Resources, SINTEF Ocean, Trondheim, Norway and Tor Nordam | Department of Physics, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, and Department of Environment and New Resources, SINTEF Ocean, Trondheim, Norway and Nicole Aberle | Department of Biology, and Centre for Autonomous Marine Operations and Systems, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and Martin Ludvigsen | Department of Marine Technology, and Centre for Autonomous Marine Operations and Systems, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and Geir Johnsen | Department of Biology, and Centre for Autonomous Marine Operations and Systems, Norwegian University of Science and Technology (NTNU), Trondheim, Norway and João Sousa | Underwater Systems and Technology Laboratory, University of Porto, Portugal and Kanna Rajan | Underwater Systems and Technology Laboratory, University of Porto, Portugal, and SIFT LLC, Minneapolis, MN, USA}, title = {Advancing Ocean Observation with an AI-Driven Mobile Robotic Explorer}, journal = {Oceanography}, year = {2020}, month = {September}, note = {
Rapid assessment and enhanced knowledge of plankton communities and their structures in the productive upper water column is of crucial importance if we are to understand the impact of the changing climate on upper ocean processes. Enabling persistent and systematic ecosystem surveillance by coupling the revolution in robotics and automation with artificial intelligence (AI) methods will improve accuracy of predictions, reduce measurement uncertainty, and accelerate methodological sampling with high spatial and temporal resolution. Further, progress in real-time robotic visual sensing and machine learning have enabled high-resolution space-time imaging, analysis, and interpretation. We describe a novel mobile robotic tool that characterizes upper water column biota by employing intelligent onboard sampling to target specific mesoplankton taxa. Although we focus on machine learning techniques, we also outline the processing pipeline that combines imaging, supervised machine learning, hydrodynamics, and AI planning. The tool we describe will accelerate the time-consuming task of analyzing “who is there” and thus advance oceanographic observation.
A light autonomous underwater vehicle is shown following phytoplankton hotspots in the coastal waters near Munkholmen in Trondheimsfjorden. In the AILARON project, the vehicle uses a CTD, an onboard camera, and an acoustic Doppler current profiler along with adequate computing power. Photo credit: Annette Stahl/Norwegian University of Science and Technology, Trondheim, Norway. > High res figure |