The ASTRAL project is developing a variety of technology to enhance IMTA, including an AI-powered water quality sensor.

EU-funded aquaculture project develops cheaper, more flexible water sensor


Researchers working on an EU-funded project supporting integrated multi-trophic aquaculture (IMTA) have developed an AI-powered sensor to automatically assess water quality.

The low-cost system harnesses artificial intelligence to model various physicochemical parameters that indicate the presence of organic matter that poses a risk to water quality in aquaculture sites.

The sensor is one of seven technologies developed by researchers contributing to ASTRAL (All Atlantic Ocean Sustainable, Profitable and Resilient Aquaculture), a €7.93 million (£6.93m), 48-month project supported by EU Horizon 2020 funding. Its core aim is to define, support and promote IMTA production in the Atlantic area.

To assess the amount of organic matter contained within a water sample, the technology measures the amount of absorbent and fluorescent spectra. That data is then passed to a planktoscope microfluidic slide, which takes images of the water sample before sending them to the cloud for further analysis.

More parameters

Currently, physicochemical parameter sensors are very expensive and limited in the number of parameters to be measured by a single sensor. By combining full absorbance and fluorescence spectra with AI-powered inferences and with automated operations, the ASTRAL technology provides a multi-parameter sensor that can be adapted to any type of water or liquid.

ASTRAL seeks to anhance IMYA at sea and on land.

Elisa Ravagnan, ASTRAL project coordinator, said: “If we are going to deliver solutions that reflect the demands of society and the environment, we’re going to need to deploy technology solutions that can quickly analyse data and detect risk.”

Ravagnan, from Norwegian independent research, science, and innovation institute Norce, added: “The electronics board embedded into the AI Water Quality Sensor is extensible and communicates with any wired or wireless network. This makes this device very adaptable to any type of water in almost any environment.

“If we succeed in communicating the potential offered by the AI Water Quality Sensor for both industry and society, we can generate tremendous impetus and drive investment.”

Work package leader Professor Marcelo Pias, from the Centre for Computational Sciences at Rio Grande Federal University (FURG) in Brazil, believes that the AI Advanced Water Quality Sensor could give the Atlantic community an innovative lead by optimising natural resource utilisation in aquaculture while also meeting the ambitions laid out in the European Green Deal.

Aquaculture Helix

The ASTRAL project brings together world-leading academics, industry experts, high-tech SMEs, future-orientated policymakers, and technological entrepreneurs and is the key project within Aquaculture Helix, part of the Crowdhelix open innovation platform that forges links between an international network of researchers and innovative companies.

Crowdhelix chief executive Michael Browne said: “The Aquaculture Helix has become a thriving community of experts and stakeholders who are positioned at key junctions throughout aquaculture’s global value chain.

“ASTRAL is indicative of the potential of the Aquaculture Helix. It brings together a team of multidisciplinary experts who collaboratively build impactful solutions such as this AI-powered sensor to automatically assess water quality.

“Building these teams can take time, which is why we support researchers and industry experts to identify potential collaborators so that they can focus on developing solutions. Our Aquaculture Helix has grown to include over 300 international experts who are eager to engage in open innovation, develop new technologies and share their results among this focused community.”