Ocean research and environmental monitoring requires water samples to be analysed to map their contents in terms of biology and minerals. In this project we will develop the technology needed for water sampling to be done automatically from unmanned surface vessels or autonomously without human intervention from manned vessels such as fish farm work-boats and transportation vessels.
The system shall be able to fix water samples so that they can be brought to shore for laboratory analysis, or analysed onboard using AI/machine-learning on data from bio-optical sensors (spectroscopy and micro-cameras). The system can take water samples anywhere from the surface down to 20 meters depth, and the decision to take water samples shall be automatic. For example, onboard analysis using bio-optics can decide if a water sample shall be taken and brought back to shore for lab analysis, since only a small number of such samples can be taken during a mission.
The project shall show the usefulness of such a system, which is expected to provide more efficient water sampling and analysis. This efficiency should be measured both in the form of costs, usefulness and environmental footprint of the operation.
The project targets a key challenge in ocean observation using autonomous systems, which is how to make in situ water sampling and real-time analysis of such samples as an autonomous process onboard small conventional vessels and autonomous surface vessels (ASVs).
The scope includes both bringing water samples back for onshore laboratory analysis, miniaturized autonomous onboard laboratory where machine learning will be used to address bio-optic data analytics, and adaptive techniques to collect samples at optimal locations and times to maximize information.
For validation of the methods, we plan to use ASVs and fish-farm workboats for environmental monitoring, including phytoplankton bloom dynamics as a case of high relevance for the society in general, and Harmful Algal Blooms (HABs) and salmon lice larvae that are particularly relevant for the aquaculture industry.
The project takes an inter-disciplinary approach to bring researchers, having expertise on enabling technology in AI/autonomy, together with researchers in chemistry and marine biology, having expertise in ocean bio-optics and water sampling, as well as industry. It is expected to not only lead to fundamental knowledge, but also to innovations that can be exploited by industry and government agencies.