Energy Harvesting for Environmental Sound Classification
Project Description:
The purpose of the project is to design a front end (pre amplifier) and innovative powering solution to interface with a high quality acoustic sensor to accurately capture the electrical signal produced by natural acoustics. It is critical to provide a high quality low noise front end in order to eventually process these signals into an identifying algorithm to be used in the natural environment for identification of species and other detection use cases. The ability to record and identify sound over a certain region gives us a geographical mapping over space that, if implemented on a large scale, allows us to track migratory patterns and draw ecological conclusions from such movement. This design must be durable to withstand a variety of environmental impacts. It must be sturdy enough to handle physical turbulence caused by environmental debris and wind, but there must be some sort of opening in order to allow sound to be recorded by the encased microphone. This opening must be covered by a material that still allows sound to pass through relatively unimpeded, but provides sufficient weatherproofing to keep out rain, dust, liquids, wind, and humidity; the frontrunner for this material Gore-Tex. We aim design a low-power high-efficiency system relative to the components available to us in order to extend the lifetime of the processor, as well as determine a battery that can provide enough charge to last for a significant amount of time while still being light enough to not make the system too heavy. For long-term implementations, we have considered the addition of a renewable energy source to continuously power the system. Powering through thermal gradient is an idea known but innovated on very scarcely. Tapping into this ability will result in open doors and ideation on the classical solution of solar cells to power environmental electronics for long durations of time.
Project Photo:
Hardware setup with Blues network