A reparameterization and monitoring platform for energy efficient GPS trackers in behavioural analysis of wildlife

In this paper, we present the architecture of a platform designed to monitor wildlife behaviour. The idea starts from the need to monitor the behavioural habits of wildlife throughout the year, in large observation areas, and with regular monitoring phases (i.e., at specific day and night times and for a certain number of months), together with the possibility of modifying the time when a device sends the localization data through a remote adaptive reparameterization process. To have a cost-effective solution in the wide areas to be covered, the platform implementation aims to monitor wildlife in an adaptive way, in relation to the seasonality and anomalous animal behaviours, obtaining reliability and energy efficiency. The platform is composed of GPS tracker devices that send localization data to a Field Programmable Gate Array (FPGA) board that collects the data and uses them for making decisions about the reparameterization of the devices. Next, the data are forwarded to a Cloud application used for the monitoring and where specific algorithms identifies the optimal strategy for remote reparameterization of the devices in case of any anomalies in the behavioural wildlife. To achieve flexibility and scalability of the architecture, modern virtualization technologies have been proposed for the development of the monitoring application using a time series database running on a Cloud platform.