I'm developing a project that combines environmental sensing with sensory translation, and I'd love to get technical input, pointers, or hear from anyone who has worked on something similar.
The idea:
I want to attach sensors to trees in a forest — likely accelerometers or piezoelectric vibration sensors — and capture the subtle vibrations trees emit: responses to wind, internal stress changes, growth movement, and possibly even the flow of water through the trunk.
In the first phase, the goal is to translate the raw data into other media — sound and/or color — by mapping the vibration signals to sonic parameters (pitch, amplitude, timbre) and dynamically shifting color palettes in real time.
In the second phase, I’d like to make the system fully live: streaming the sensor data from the forest to a separate space — a gallery, installation room, or studio — where it is translated in real time into sound and/or light, allowing visitors to experience the “presence” of the forest remotely.
Questions I’m currently working through:
Which sensors are best suited for capturing tree vibrations in outdoor field conditions?
What are established or interesting approaches for mapping vibration data to sonic parameters?
Which communication protocols would you recommend for live sensor-data streaming from a field environment (Wi-Fi / LoRa / LTE / something else)?
Thanks in advance — I’d really appreciate any references, ideas, or examples of related work.
One at a time. Accelerometers, geophones and piezo disks respond to different ranges of vibration intensities and frequencies, and require different circuitry.
It is a good idea to decide whether you can actually measure something interesting, before trying to figure how to send the data elsewhere.
Perhaps you should decide on the ranges of frequencies you wish to sense. In a forest you will have frequencies from zero to supersonic. A single sensor will not be able to convert that full range into an an electrical signal for your Arduino.
For color my library might be useful as maps float variables e.g. from sensor to colors.
It can interpolate between (RGB) color points, but also allows "sharp edges"/
For each possible "influence" to the sensor, also have a "control" sensor. For example, have "wind" sensors (rate and direction) near, but not attached to the tree. Same with a rain sensor (current and accumulated), Sunlight (intensity over time). Trailcam (what animals are chewing on your experiment, is a bear climbing the subject, are bees nesting inside the tree).. so much fun stuff. Otherwise, you get ducks.
For a professional example, consider this 2D/3D tomography system that uses vibration sensors to map and image rot inside the trunk of a living tree. The user supplies the vibrations with a small hammer.
Does the ARBOTOM system require active input — meaning the user must physically strike the trunk with the hammer each time in order to acquire data? Or can it also operate passively, capturing ambient vibrations on its own?
A value (measurement) is of no use without a control (measurement). What is "100?" If the control is "100" then the value is little. If the control is "10" then the value is large.
There are wild camera's that are attached to trees by means of a belt with some soft layer.
These do not harm the trees.
Do not use tie wraps as these may damage the tree, esp. when you remove them.
Just order them and start prototyping just to learn how such sensor works.
Put the sensor and Arduino in a box with an SD card module and take it for a walk.
You should be able to see the vibrations of your footsteps.
That way you learn first hand the range and sensitivity of the sensors.
I don't understand - it will sense my vibration while working? (when I hold it with my hand)
how it can sense the tree vibration?
I have real hard time to understand how this type of sensor can sense tree vibration
