Scientists across Europe will revolutionise how we measure and monitor our changing oceans in a new 8.9 million research programme.
The TechOceanS project, coordinated by the National Oceanography Centre, will deliver a new generation of low-cost and easily deployable sensors capable of wide-scale monitoring of ocean bio-geochemistry.
The international programme will pioneer five new sensors, two imaging systems, a novel sampler and an Artificial Intelligence-driven image processing methodology, all capable of operating at depths beyond 2,000 metres.
Southampton experts in Electronics and Computer Science, including Professor Hywel Morgan and Dr Daniel Spencer, will use their skills in microfluidics and sensor design to develop miniature, autonomous biological analysers capable of continuous monitoring and characterisation of marine organisms (such as harmful alga) and micro-plastics in the oceans for many months at a time.
The interdisciplinary research will include Southampton's Dr Peter Glynne Jones and his expertise in the manipulation of microscopic particles using the forces generated by ultrasonic waves.
Professor Morgan, Head of the Biomedical Electronics Research Group, says: "The oceans are essential in regulating climate and produce most of the planets oxygen, yet are ruthlessly exploited. To better understand the health of the oceans we urgently require significantly improved ways of measuring key chemical, biological and ecosystem variables.
"This ambitious and exciting project led by the world-class National Oceanography Centre brings together internationally leading partners from all over Europe. It will revolutionise ocean measurements and will have long-term impact on our understanding of life in the oceans and the way we exploit these resources."
TechOceanS will actively engage with scientists, manufacturers, marine stakeholders and resource managers to target Essential Ocean Variables that currently can only be measured by observations from a ship.
Engineers at Southampton will develop automated methods to turn vast numbers of images taken in the ocean into numerical information about the distribution of species, habitats and plastic pollution, and transmit this in near real-time over the low communication bandwidths available in the open ocean.
This will allow researchers to be instantly aware of discoveries made by autonomous systems, regardless of where they are in the world, and allow people to adapt data acquisition strategies based on up-to-date information.
Professor Blair Thornton, Director of the Centre of Excellence for In Situ and Remote Intelligent Sensing (IRIS), says: "TechOceanS will be revolutionary for how imagery is used in ocean science. Most ocean sensors make continuous measurements of scalar values like temperature, salinity or pH, where the size of data is small enough to transmit over the acoustic or global satellite bandwidths available in the open ocean. Images on the other hand, are far too large to transmit directly, and the require non-trivial analysis to extract useful information from the scenes they capture.
"This means that people normally need to go out to sea and physically recover imaging systems to make the data accessible. For remote sites that are rarely visited, this can introduce delays of months or even years from when the original events took place.
"The edge computing methods we are developing in this project will make the cameras smart enough to reliably extract and summarise important information by themselves, and through intelligent interaction with remote hosts, transmit these finding efficiently and in near real-time. R