Would there be interest in other sensors than the current ones, like sound pressure level sensor or light intensity? Would imagine that they wouldn’t be too difficult to implement on the current beacon board.
An audio sensor could be used for applications like engine monitoring, burglar alarm, wildlife monitoring etc. A light sensor could be used in similar applications but also as a trigger for lights in places where you don’t want to put cables for a sensor.
Both sensors would potentially carry private information, so if there is not yet the option available, then we really should have the option for encc ftrypting the beacon message payload.
I have some expertise on the audio side and could chip in if someone picks up the project. Through my work I have some further resources available for both audio calibration and encryption, but for hardware design I have zero experience.
We’re actually looking into photodiodes for light intensity detection right now. Would you have some specific audio chip in mind? nRF52 supports I2S in case that is necessary, in general low power consumption or really fast start/stop of audio chipo would be a must.
nRF52 has AES-128 encryption HW built, in and we use it in some special cases. Key could be for example the device ID which can be read via NFC.
I don’t know anything about hardware so I don’t know relevant chips. For sound pressure level measurements you’d get best accuracy with continuous measurement with energy accumulation, but I guess a short segment like 30ms would be passable as well. If it’s just a short segment, then a generic audio chip with built-in mems microphone could be simplest.
If it’s a generic audio chip/microphone, then it would be possible to do some fancy speech activity detection on chip as well. The 45 MIPS speed would not be a problem, though the amount of available 64kb RAM might limit performance. (Hope I got the numbers right).
PS. Implementation would be a nice Master’s thesis project if someone here is interested.
I am interested in detecting and recording short periods of sound (maybe ten second clips 15mins apart) in beehives in the 100 to 5000 hertz range at very low levels with an add-on device to the Ruuvitag (that could have its own battery source) and then send the data over a cellular link to read the files with spectrum analysis on a PC or Mac.
Excuse me if my language is non-scientific.
