Detecting Illegal Logging Using An Acoustic Monitoring System
Ghana has a strong reputation for innovation in timber processing, making products particularly suited to the European market. Around 11% of Ghana’s timber exports are sent to the EU according to an EU FLEGT report. The forest sector is the fourth largest contributor to Ghana’s GDP.
The same report states that Ghana has around 2.6 million hectares of forest reserves dedicated to timber production and an additional 2 million hectares of cropland that also produce timber. It further highlights that the country also has 500,000 hectares of unreserved forests.
However, the country is currently battling illegal logging or deforestation as the act has resulted in more than 70% reduction of forest trees. The significance of forests has been preached severally, and as such should not be belabored. We are well aware that forests are key producers of oxygen, contribute majorly as air filters, and provide food security.
To put an end to deforestation and illegal logging, Concept-X, a team from the Hacklab Hackathon, is set to launch the DAMS Project (Deforestation Alert and Monitoring System).
Founded by Eugene Sewor, Tuglo Emmanuella, and Austin Amegbe, the DAMS project is an initiative that aims to report illegal loggers to the authorities and to support advocacy for forests.
The DAMS project is currently a bootstrapped project that uses an artificial intelligence system to detect sounds from chain saws and other machines used in forest logging and transmits feedback to authorities. The web-based system has been trained to differentiate between chainsaw sounds and other machinery.
The project will begin with deploying the system in forest areas with a connection to authorities. Once an activity is detected, alerts will be collected in a database managed by the authorities, the local government, or other stakeholders. Data displayed in the authorities’ database will illustrate a particular hotspot where there is a surge of illegal logging activities within a particular time frame.
Moving forward, the team will develop the web-based system into a deployable device that could help create a more robust monitoring system.
“We have trained the system with a chainsaw, motor, and other machinery sounds, and our system quickly detected the difference. We will develop the system into a more deployable device as that will perform a better inference and enable a quick transmission of alerts,” said Austin.
Furthermore, Concept X plans to combine forest monitory with sensitization programs on environmental sustainability. The team will organize afforestation campaigns and lead locals in planting exercises. They estimate that they can lead people to grow more trees rather than cut them down.