| Resumo: | The observation and treatment of animals in veterinary hospitals is still very dependent on manual procedures, including the collection of vital signs (temperature, heart rate, respiratory rate and blood pressure). These manual procedures are time-consuming and invasive, affecting the animal’s well-being. In this work, we purpose the use of IoT technologies to monitor animals in hospitalization, wearing sensors to collect vitals, and low-cost hardware to forward them into a cloud backend that analyses and stores data. The history of observed vitals and alarms can be accessed in the web, included in the Pet Universal software suite. The overall architecture follows a stream processing approach, using telemetry protocols to transport data, and Apache Kafka Streams to analyse streams and trigger alarms on potential hazard conditions. The system was fully implemented, although with laboratory sensors to emulate the smart devices to be worn by the animals. We were able to implement a data gathering and processing pipeline and integrate with the existing clinical management information system. The proposed solution can offer a practical way for long-term monitoring and detect abnormal values of temperature and heart rate in hospitalized animals, taking into consideration the characteristics of the monitored individual (species and state).
|
|---|