BUDGET EXPENDITURE: $87,538

PRINCIPLE INVESTIGATOR: Dr. Andrew Trotter

ORGANISATION: University of Tasmania (UTAS)

PROJECT CODE: 2023-035

PROJECT STATUS: Current

 

WHY IS THE RESEARCH BEING UNDERTAKEN?

Traditional methods of obtaining oyster biometrics rely on manual measurements, which are not only labour intensive but also time-consuming and costly. This constraint will often restrict the approach in data collection 
in both farming and selective breeding settings, due to cost and logistical limitations. OysterMetrics, an early prototype AI tool, offers the potential to greatly simplify and accelerate the data collection process by utilising cutting-edge computer vision and machine learning technologies. 

With far greater ease in data collection, producers will be able to test various aspects of farm management 
practices, analysing trends and patterns in oyster growth, leading to data-driven decision making. Such 
assessments could be based on either routine farm operations or via designated trials to test the effect of 
different environmental of farming variables. This should lead to improve yields and overall farm productivity 
and contribute to the growth of the industry. 

Likewise, in selective breeding, data collection remains a limitation in maximizing the gains from the 
investment that has been made. Using AI for data collection will lead to more accuracy in the data collection 
in term of measurement accuracy and errors associated with small subsamples. More data leads to greater 
statistical power, meaning that the results of the analysis are more reliable and robust, and also eliminates 
user bias. Our work on the prototype found manual width measures performed by three people could vary by 
up to 8 mm (on approx. 30 mm wide oysters). With more data points, breeders can make more confident 
conclusions about the relationships between different traits and the genetic factors influencing oyster growth. 
This has significant potential to deliver faster genetic gains in growth, uniformity and shell shape, which would 
improve farm productivity. 

​

OBJECTIVES:

1. Build four opertaional Oystermetrics units.

2. Test  four operational OysterMetrics units in both laboratory and field settings at IMAS and in the field 
   under normal ASI selective breeding operations. 

3. Provide three operational OysterMetrics units for industry evaluation (one unit to be evaluated in each 
   of SA, TAS and NSW).

4. Refine the operation of the OysterMetrics tool based on field and laboratory testing, and industry 
   evaluation and deliver the final operational versions to industry. 

​

​PROJECT RESOURCES GENERATED:

• Link to project page on the FRDC website

• No project resources / outputs are currently available.