Introduction

The patent proposes a system for the localization of tortoises during the phases of nesting. The system builds upon a localization system (GPS), a wireless communication system  and on an algorithm based on efficient neural networks models for the analysis of sensor data such as accelerometers, light and temperature

Technical features

Currently there are no devices able to detect the eggs deposition phase of tortoises, and consequently to transmit the position of nests of wild tortoises in their natural environment. Our system processes the measurements of different sensors inside a device attached to the carapace of female turtles, to recognize the deposition phase and to transmit the coordinates of the nest for the intervention of the safeguard operators of the species. The algorithm for the recognition of the deposition pattern uses efficient automatic learning techniques (small-scale neural networks) that allow an accurate and flexible recognition of such sensor patterns in different context and for different species of tortoises. The pattern recognition system  has been found to be extremely effective in accuracy on experimental data collected in a campaign of real cases of spawning turtles.

Possible Applications

• Automatic monitoring of wild animals for scientific purposes;
• Programmes to protect and safeguard wild turtles.

Advantages

• Automatic identification of the excavation activity of a turtle’s nest (related to spawning);
• Automatic localization of wild turtle nests;
• Continuous monitoring and habitat of wild turtles for long periods.