Towards A Low-Cost Localization System For Small Ruminants

Monitoring the activity and location of weeding flocks bears distinctive benefits to the livestock industry. Among other benefits, access to this kind of data allows farmers to manage pastures more effectively, monitor the animal's reproductive cycles and health condition, as well as detect threats from external agents, such as attacks from predators or thieves.The scarcity of human shepherds, together with the significant and increasing costs associated with such activity, fostered the research of ICT technologies as a means to automate the collection of location data. GPS trackers are the most popular solution, but they are expensive and have high power consumption, thus impairing its use for small ruminant flocks. Consequently, alternative solutions, such as the ones based on advanced RangeRSSI-based algorithms are potentially more attractive, especially when a communication infrastructure already exists. This is the strategy followed in the scope of the SheepIT project, whose localization mechanism is presented in detail in this paper. Together with the localization mechanism, the farm radio infrastructure allows tracking animals in real-time. To evaluate the localization system performance, this paper compares the results of the implemented framework with the results obtained when using a commercial GPS tracker. Although GPS allows obtaining a better accuracy, the RSSI-based mechanism herein proposed allows, in certain conditions, attaining accuracy levels that meet the requirements of this type of application, without requiring additional hardware and with residual power consumption.