Why use bioacoustics?

Bioacoustics is valuable for biodiversity monitoring because it is non-invasive, allowing for continuous, long-term data collection without disturbing wildlife. It is especially effective for detecting cryptic and nocturnal species, and it can monitor a wide range of taxa. The method is efficient, scalable, and well-suited for remote or difficult-to-access areas. With advancements in AI and machine learning, large datasets can be analysed automatically, making it easier to track biodiversity trends.

Why use the bioacoustics service from Kumbe?

At Kumbe, we offer our bioacoustics services alongside other methods like eDNA and camera traps. You have the option to use your own sound sensors or lease them from us for the duration of your project.

Based on your objectives, the sound sensors are strategically placed, often in conjunction with eDNA sampling and camera traps. While Artificial Intelligence for sound analysis is still being developed, and specific AI models for Tanzanian habitats are not yet available, Kumbe will utilize acoustic indexes. These indexes help estimate the number of species in an area, and by monitoring over time, they can indicate whether biodiversity is increasing or declining. Additionally, our local experts carefully annotate the collected sound samples, which can later be used to train AI models for Tanzanian ecosystems.

Kumbe’s service for bioacoustics is currently in development. If you like to join our pilots, please contact us!

Which services does Kumbe offer?

Plan

Based on your specific requirements and objectives, Kumbe's team of experts will develop a tailored sound sensor monitoring plan. This plan will include the following components:

  • Where to place the sensors

  • How many sensors to place

  • Which settings to use on the sound sensors

  • The duration of the placements

Kumbe works with local communities to place the sound sensors at the predefined locations.

Annotation

Once the SD cards are collected, there are 3 approaches possible to process the data:

  • manual annotation of all sound samples

  • use acoustic indexes to understand the number of species

  • store the samples and process them, once better AI models are available

Analysis and reporting

Building on your initial requirements, we will conduct a thorough analysis of the results, delivering comprehensive and insightful reports. These reports will include detailed data interpretations, relevant visualizations, and actionable recommendations tailored to your specific needs, ensuring clarity and informed decision-making