SERKET is an innovative and disruptive Animal Recognition Software able to collect data using standard security cameras (already available in the livestock), analyze the data by a proprietary Artificial Intelligence algorithm that detects behavioral pattern anomalies and automatically define the best strategy for the livestock health management. SERKET uses deep learning technology to detect relevant animals in images and videos, keep track of the performance parameters of animals and hence monitor their health: starting with the huge amount of annotated data and trained neural networks, the models are used to produce inference able to perform animal detection and health monitoring. Basically, SERKET uses anomaly detection to identify the sick animal based on its abnormal behavior and support the farmers in the implementation of the most reliable, non-invasive and cost-competitive solutions.
The main benefits are:
1. Reduction of antibiotics usage - instead of treating all the animals with drugs and antibiotics, SERKET can early detect the ailments and focus only on the animals that need the treatment. With fewer antibiotics and drugs usage, the risk of the occurrence of Antimicrobial resistance (AMR) for the animals (and also for human) is prevented. Moreover, the reduction of antibiotics usage leads to a significant saving for farmers.
2. Reducing the mortality rate - according to Wageningen University and other health studies, SERKET solution can lessen the livestock mortality rate thanks to the deep learning based behaviour recognition system. This leads to a reduction in value losses and more sustainable solution to the farmers.
3. Support on livestock management - SERKET is capable of counting animals, of determining feed requirements and automatically adjusting liquid feeding systems.
4. Fulfill and prove of the compliance of the animal welfare conditions – with the identification of the harmful and aggressive animal behavior (for example ear biting and tail chewing) in the pens, SERKET supports the farm audits.
