Development of a computer vision-based device for continuous larval monitoring in commercial-scale crustacean aquaculture

Title : Development of a computer vision-based device for continuous larval monitoring in commercial-scale crustacean aquaculture

Abstract:

This study presents the development of a novel computer vision-based device for larval monitoring in large-scale crustacean aquaculture. The device utilizes computer vision algorithms to identify and count larvae within a grow-out tank, providing continuous and real-time data for estimating the total tank population. During a larval grow-out trial, the device underwent thorough testing to validate its functionality. Shortterm sampling revealed distinctive larval behaviours triggered by environmental conditions such as lighting variation, or the presence of feed, as well as circadian patterns. Furthermore, longer-term measurements demonstrated the device's ability to accurately track population levels over time. The implementation of this device holds significant commercial potential in the field of crustacean aquaculture. The continuous and real-time data stream generated by the device enables improved run management, reduced labour requirements, optimized feeding practices, and precise scheduling of downstream operations based on up-to-date projections of batch output. Through real-time larval monitoring capabilities, the developed device offers valuable insights into larval population dynamics, enabling aquaculture managers to make informed decisions, enhancing overall production outcomes and profitability.

Audience Take Away Notes:
• How the automation of manual processes using computer vision and machine learning can improve operational efficiencies and reducing labor costs.
• How real-time data provided by the presented device enables data-driven decision making for optimizing feeding practices, scheduling operations, and improving run management.
• How the device streamlines data collection, offering cost-effective insights that would be time-consuming and costly to obtain manually.
• The presentation showcases the potential and benefits of automation in the crustacean aquaculture industry, and the aquaculture industry at large, emphasizing streamlined processes, data-driven decision making, and technology adoption for improved operational outcomes.

Biography:

Dr Dean Giosio graduated with a combined BEng(Hons, 1st class)/BSc(Chem) from the University of Tasmania in 2009. Following a PhD in fluid dynamics from the same institution, he then held the position of postdoctoral research fellow at the Australian Maritime College Cavitation Research Laboratory. Here Dr Giosio worked on a collaborative project with the United States Office of Naval Research investigating microbubble populations and cavitating flows about hydrofoils. Since 2018 Dr Giosio has been working to scale-up systems and commercialize technology developed at the IMAS ARC Industrial Transformation Research Hub for Sustainable Onshore Lobster Aquaculture.