A multidisciplinary team of researchers from the Peruvian Union University (Universidad Peruana Unión, UPeU), Tarapoto campus, has developed a high-impact prototype for the Amazonian aquaculture sector. It is an intelligent device that seeks to optimize the sex reversal process in tilapia fry, a key technique for producing this species without resorting to invasive methods.

This project, called Intelligent device for the technical improvement of the sex reversal process in tilapia fry based on PCR (Polymerase Chain Reaction) and convolutional neural networks, proposes an innovative alternative to identify the sex of fry with greater precision and speed, without the need to sacrifice samples, as currently occurs with traditional methods, and using more efficient and sustainable technologies.

An Innovative Technological Solution

Tilapia is one of the most farmed species in the world due to its high demand. In the San Martín region of Peru, its production represents an important part of the local economy. However, the traditional method for determining the sex of fry, the initial life stage of the fish, involves the slaughter of at least 100 fish per thousand, and can take two to three weeks to complete. Furthermore, this type of procedure, based on the analysis of histological slides, is expensive, slow, and unsustainable.

Faced with this situation, the team of researchers from the Professional School of Systems Engineering of the Faculty of Engineering and Architecture of UPeU developed a prototype based on two key technologies:

• Polymerase chain reaction (PCR):  allows early and accurate identification of the sex of fry by amplifying specific DNA sequences. 

• Convolutional neural networks (CNNs):  These use artificial intelligence algorithms that, through computer vision, recognize the physiological characteristics of fry at six weeks of age, streamlining the process without harming them.

Sustainable Project

According to Yngue Ramírez, a master's degree holder in engineering and one of the principal researchers, "this system will prevent these losses, as it will allow for accurate results from the fish's sixth week of life, using specialized cameras and artificial intelligence. Furthermore, analysis time will be significantly optimized, as identification can be completed in just 45 seconds."

With a 60% completion rate, this innovative device promises to be commercially available nationwide once the research and validation process is complete.

Research Team

The team works in partnership with the Ahuashiyacu Aquaculture CITE, an institution based in Tarapoto, Peru, that provides technological services to the aquaculture production chain. The team is validating the prototype and addressing real needs of the sector. During a technical visit to the CITE, the researchers identified technical challenges related to the fingerling sexing process that could only be resolved through technological innovation.

The original article was published on the South American Division Spanish news site. Join the ANN WhatsApp Channel for the latest Adventist news.