Kai Cursons

Kai Cursons


I graduated from Cornwall College Newquay (in partnership with Plymouth University) in 2019 with a BSc (Hons) in Applied Marine Zoology. I have since been studying for an MRes in marine biology at the Marine Biological Association of the UK (again in partnership with Plymouth University).

During my undergraduate study, I developed an interest in the application of physical and digital technologies in marine science. In particular, how advances in technology may be used to supplement existing resources or offset deficits where resource limitations exist.

During both my undergraduate and postgraduate studies I have cultivated this interest through application in invasive species management. During my undergraduate degree, I utilised regression-based Machine Learning (ML) to model habitat suitability for the Pacific oyster (Magallana gigas) in southwest England. For my MRes, I have expanded on this through further exploration of ML and AI to create an improved distribution model for M. gigas. This model increased spatial extent to cover the whole UK and additionally incorporated climate forecast data to explore how population dynamics may change under various future climate scenarios.

PhD title: Augmented taxonomic analysis of Continuous Plankton Recorder data using Artificial Intelligence

Having recently celebrated its 90th anniversary, the Continuous Plankton Recorder (CPR) survey represents the longest running and most geographically extensive plankton survey in the world. The CPR itself is a torpedo shaped plankton sampling device designed to be towed behind ships, filtering plankton out of the water column onto a moving band of silk.

For over 61 years survey protocols have involved the manual taxonomic identification and counting of plankton samples enmeshed in the silk through human analysts via microscope. Recent advances in technology have now enabled the incorporation of sensing and imaging systems into the CPR itself increasing observational capabilities.

The aim of this PhD is to process plankton images from a range of ex situ and in situ sources to develop a new generation of merged digital image libraries. These libraries will then be used to train AI models (e.g. Convolutional Neural Networks) to automatically identify and classify plankton images through transfer learning. This should result in the ability to analyse samples in real-time, augmenting the sample processing capabilities of the survey.

Awards and Prizes

Cornwall College Group student of the year award

Other information

Project officer with the Community Invasive Non-Native Group (CINNG)