Illegal logging is a major issue worldwide, and it is estimated that over 100 million cubic meters of forest are logged illegally each year. This exacerbates not only localized environmental degradation but also contributes a large portion of global carbon emissions (approximately 20%). In 2017, six researchers based in Sri Lanka approached the tricky issue of how to monitor illegal logging in real-time using Internet of Things (IoT) technology, ecoacoustics, and artificial intelligence; the result is a new system of monitoring called TreeSpirit. Listening posts, small sensors fitted with IoT sensors and ecoacoustic sensors, transmit audio information to a sound processor node in the forest, with one node supporting three posts in the forest. The node uses a complex neural network, a form of artificial intelligence, to pick out the sound of chainsaws from the audio. If a chainsaw is detected, the audio is sent to the Base Station Computer, which receives audio information from the nodes, which then sends an alert via the Cloud to a website or generates an SMS text message to authorities. Because the nodes support just three listening posts with a distinct position, authorities can get an approximate location of where the illegal logging is occurring. Being able to pinpoint the exact location would be optimal but getting the exact location from audio data was not achievable. Considering illegal logging is difficult to stop when it is happening, this research represents an interesting leap forward in being able to monitor and stop it in real-time. 


Artificial Intelligence, Illegal Resource Extraction, Monitoring, Regulation