Human-Wildlife Conflict Early Warning System Using the Internet of Things and Short Message Service.

Human-wildlife conflict (HWC) is an important challenge to communities living in areas bordering wildlife game parks and reserves. It is more evident in the United Republic of Tanzania, whose economy depends on wildlife tourism. This paper proposes a low-cost and low-power early warning system using the Internet of Things (IoT) and Short Message Service (SMS) to support HWC respond teams in mitigating these challenges. The system comprises three primary units: sensing, processing, and alerting. The sensing unit consists of a Passive Infrared (PIR) sensor, a Global Positioning System (GPS), and a Raspberry Pi camera. The PIR sensor detects the proximity of the animal using the heat signature, GPS senses and records the current location, while the Raspberry Pi camera has the primary purpose of taking a picture after the PIR sensor detects the proximity of the animal. The processing unit with a Raspberry microcomputer performs data processing and image inferencing using the You Only Look Once (YOLO) algorithm. Last is the alerting unit, which includes a Global System for Mobile (GSM) communications module for sending SMS messages to the human-wildlife conflict response team and the nearer community response team leader whenever wild animals are spotted near the park’s border. The system detects, identifies, and reports the detected wild animals. The GPRS provides internet connectivity to support data collection, storage, and monitoring in the cloud. [ABSTRACT FROM AUTHOR]