Kenyan scientists use AI to track bird songs and monitor forest health
In a groundbreaking initiative, researchers at Dedan Kimathi University of Technology’s Data Science and Artificial Intelligence Laboratory (DSAIL) are deploying smart bioacoustic devices to monitor Kenya’s forest ecosystems through bird songs.
The solar-powered devices, equipped with sensitive microphones, record bird vocalizations during peak activity hours – from 5 AM to 10 AM and 5 PM to 7 PM.
These recordings are then converted into spectrograms and analyzed using artificial intelligence to identify different bird species.
« When an acoustic event is detected, the microphone captures it, and the data is stored for analysis, » explains Jason Njoroge, a junior researcher at DSAIL.
The project aims to assess ecosystem health by tracking bird populations, which serve as crucial indicators of environmental change.
According to Samuel Guchu, an ornithologist at the Kenya Bird Map project, this research is particularly timely. « Birds are excellent indicators of ecosystem health. Their ability to quickly relocate when conditions become unfavourable makes them perfect early warning systems for environmental degradation, » he notes.
Current projections suggest that vulnerable bird species could lose 20-30% of their habitat in the next decade due to human activity and urbanization.
The project complements other conservation initiatives like the Kenya Bird Map, which tracks bird distribution patterns across the country.
By combining AI technology with traditional ornithology, researchers hope to better understand and protect Kenya’s diverse forest ecosystems and their avian inhabitants.
Source: Africanews
