PrawnAugmented reality in aquaculture: helping Australian prawn farmers move to the forefront of efficient, high production, low-risk prawn aquaculture through near real-time sensing and forecasting of animals and ponds

What problem could our augmented reality tech help solve in the aquaculture industry?

The worldwide demand for marine-based protein is booming. Aquaculture is an intensive farming system capable of returning high yields of protein for human consumption and plays an important role in food security in many countries. Australia’s prawn industry produces in the order of 5000 tonnes of prawns annually, and this is set to expand. Just 1% production improvement in farming practices could bring a further $16 million to the Australian industry annually.

Aquaculture is undergoing a transformation in scale both in Australia and globally, however there’s a substantial yield gap between how much is harvested and how much could be harvested.

What did we do to help solve the problem?

At Digiscape, we worked on novel techniques and technologies in sensing, data modelling, situational awareness and decision support to help the industry transform to more consistent yields, across far greater scales, to capture unrealised profits.

A man with a heads up display

Digiscape’s Dr Mingze Xi showing off one of the wearable technology systems we worked on

Wearable Data Collection Suit is a smart glass application that we aim will transform how aquaculture farmers carry out their field jobs via hands-free interaction, situated water trend visualisation, automatic sensor data extraction and deep learning-enabled crop sampling. We hope the system will play a vital role in upscaling the prawn farming industry by supporting staff to make more informed farm management decisions, faster than they can currently.

We also created a decision support application for prawn farmers that makes use of real-time sensor data and state-of-the-art deep learning models to provide 24-hr forecasting and anomaly detection. Water quality in ponds is highly dynamic and challenging to manage for large farms with hundreds of ponds.

Our application provides farmers with valuable information that allows them to be more proactive with pond management, thereby improving water quality and reducing risk. The application has been tested on a prominent Australian prawn farm, who have provided highly promising feedback.

Where can I find more information?

Read our media release or blog post.

Read select publications:

He WN, Xi, MZ, Gardner H, Swift B, Adcock M (2021) Spatial anchor based indoor asset tracking. pp 255-259 in ‘IEEE Conference on Virtual Reality and 3D User Interfaces (VR 2021)’, Lisbon, 27 March-2 April 2021.

Rana M, Rahman A, Dabrowski JJ, Arnold S, McCulloch J, Pais B (2021) Machine learning approach to investigate the influence of water quality on aquatic livestock in freshwater ponds. Biosystems Engineering 208, 164-175.

Dabrowski JJ, Rahman A, Pagendam D, George A (2020) Enforcing mean reversion in state space models for prawn pond water quality forecasting. Computers and Electronics in Agriculture 168, 105120.

Dabrowski JJ, Zhang YF, Rahman A (2020) ForecastNet: a time-variant deep feed-forward neural network architecture for multi-step-ahead time-series forecasting. pp 579-591 in Yang H, Pasupa K, Leung ACS, Kwok JT, Chan JH, King I (eds) ‘Neural Information Processing (ICONIP 2020)’. Lecture Notes in Co­mputer Science, vol 12534. (Springer: Cham).

Rahman A, Dabrowski JJ, McCulloch J (2020) Dissolved oxygen prediction in prawn ponds from a group of one step predictions. Information Processing in Agriculture 7, 307-317.

Dabrowski JJ, Rahman A, George A, Arnold S, McCulloch J (2018) State space models for forecasting water quality variables: an application in aquaculture prawn farming. In KDD ’18: Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, London, August 2018, pp. 177-185. doi:10.1145/3219819.3219841

Xi MZ, Adcock M, McCulloch J (2018) Future agriculture farm management using augmented reality. In: ‘IEEE Virtual Reality 2018’, Reutlingen, Germany, March 2018. doi:10.1109/VAR4GOOD.2018.8576887

Contact a team member:

  • John has a robotics and sensing background which has been applied loosely in the marine space for a decade. His current focus is on bringing IoT and machine learning to the aquaculture and, more broadly, agriculture sectors.
  • Mingze led the augmented reality component of the aquaculture project. He is a senior experimental scientist with CSIRO's Data61.
Joel Dabrowski
  • Joel led the water quality component of Digiscape's aquaculture project. His key research areas are machine learning and probabilistic modelling in time series problems.