News

Umitron makes new waves with Remora by integrating AI-based feeding optimisation and mortality estimation to large-scale farms

Photo of author

By Milthon Lujan

UMITRON PTE. LTD. (Singapore, Co-founder/Managing director Masahiko Yamada, hereinafter Umitron) is proud to announce the launch of its latest product, Umitron Remora.

remora setup

This plug-and-play software brings the company’s highly specialised feed optimisation algorithms, pellet detection and mortality estimation to large-scale farm operations without installing additional hardware equipment. It was specifically designed to be a tool to complement a farm’s operations to increase efficiency and reduce unnecessary workload and stress for feed operators when having to monitor multiple concurrent pen feeding.

Since the Fish Appetite Index’s (FAI) launch in 2019, it has been helping farmers to gauge fish appetite using a combination of machine learning algorithms and image analytics. FAI’s integration with Remora now expands its use to a broader range of farming operations and species currently being farmed on a larger scale.

Producers can now easily access FAI by installing Remora on their computers and connecting to existing on-site cameras. Doing so allows them to track fish feeding activity in real-time and make necessary changes to feeding protocols. Using customised pre-set thresholds, operators will also receive an alert when changes are detected, helping them to track critical shifts in fish response during feeding. In addition to accessing and using existing cameras, Remora also does not require a robust Internet environment nor the installation of an additional localised server. Instead, Remora uses Umitron’s unique Edge AI technology to perform FAI analyses utilising the local computing resources of the producer’s IT systems.

Remora also boasts two additional key features – pellet detection and mortality estimation. The former goes hand-in-hand with appetite detection to provide producers with a comprehensive overview of real-time feeding operations. In this way, Remora ensures that fish are fed based on their hunger levels and that there is no unnecessary feed wastage that may contribute to unsuitable environmental conditions.

See also  BVA calls for better understanding of fish welfare and impact of aquaculture on the environment

The mortality estimation feature allows farm operators to streamline their operations by automating the daily dead fish counting process. Users start by scheduling their preferred counting times and ensuring their cameras are correctly positioned. Remora then automatically quantifies dead fish for the captured images using AI. This feature will help users manage mortality counts and check fish health conditions more easily.

All feed and mortality data are stored in the dashboard. Managers can easily access the historical data to keep an eye on feed performance and compare it with their production metrics over time. They can also use the data for training purposes to ensure consistency between different feed operators.

The company is currently working with several partners in key salmon producing regions and producers who farm other species using centralised feeding systems and is also looking for other producers keen on further optimising and refining their feeding protocols to achieve a more sustainable production footprint.

About Umitron

UMITRON is a Singapore and Japan-based deep tech company aiming to solve worldwide food and environmental problems by empowering aquaculture with technology to achieve a more sustainable footprint. Over the past few years, we’ve developed products exclusively for the industry using a combination of IoT, satellite remote sensing and AI technology.

We are committed to achieving industry growth by improving the working environment and ensuring a safe and stable supply of marine resources with a strong focus on marine conservation and protection. Ultimately, we aim to realise our mission to ‘install Sustainable Aquaculture on Earth’.

Umitron website: https://umitron.com/en/index.html

Leave a Comment