SETUP HELP WITH YOUR SENSOR?
Do you need help with your sensor? The setup and configuration page is at https://domesticlight.art/setup/
Guidance on doing your own software update is on our GitHub code page and this step-by step guide https://domesticlight.art/software-update-aug-12-2023/
OPEN SOURCE CODE:
The project code is being developed as open source code is available for re-use and review. The initial Domestic Light sensor client code for the prototype and artist proof versions of the sensors is now available on github along with our internal code reference and troubleshooting details. https://github.com/thirtysevennorth/domesticlight_public
We welcome help from programmers with experience in working with the ESP-32 S3 platform using Arduino IDE and environmental sensors and also with experience in applying machine learning to investigate whether spectral profiles from the AMS sensor can be used to help identify specific VOC gases for the next phase of the project.
If you are interested in contributing please get in touch!
WHAT DATA ARE WE COLLECTING:
The primary data gathered by Domestic Light is the color of differing spectra of light in domestic settings (windowsills) in order to build a nuanced multi-spectral data time-lapse of the color ranges of light in domestic settings world-wide in order to answer the primary question of “what is the color of home?”
The data gathered by Domestic Light centers on the output of a small sensor “bauble” that hosts of the project place on their windowsill for a year. The sensor is an AMS7341 or AMS7343 multispectral light sensor gathering the intensity of light spectra from 350nm to 1000nm.
Data collected by the sensors consists of a time-series set of data that consists packets recorded once per second that include (11) 10 bit readings of the spectral intensity of 9 light spectra ranging in wavelength between approximately 350 nanometers to 1000 nanometers, and two additional channels recording overall light intensity and flicker hertz rate if any.
Additional control and configuration metadata transmitted in each packet: a time stamp generated by the real time clock chip, a network clock timestamp, a hashed value encoding the MAC address of the microcontroller for identification, hashed key and channel information for the MQTT database and the current sensor health state. This data is recorded locally on the controller at once per second, cached by the microcontroller and transmitted using an SSL encrypted MQTT protocol to a cloud based MQTT broker SQL database (between once per minute and once per day depending on local network bandwidth.
On the hardware side: The sensor package consists of custom PCB to mount the AS7341 or AS7343 sensor from AMS, alongside an onboard real time clock (RTC) (Maxim 31343) synced at time of shipment to the other sensors, and a microcontroller (Rev A uses a Raspberry Pi Zero 2 W, Rev B uses ESP32-S3 Feather). Power is provided by a 5v USB-C power input and a power cord appropriate to the location hosted.
WHERE IS IT BEING COLLECTED?:
The data from sensors will be stored on the devices while being buffered and then transmitted to longer term cloud storage. After trasnmission the data will be stored in a virtual private server hosted by one of a number of private server providers including AWS (Amazon Web Services, Inc.), Digital Ocean (DigitalOcean Holdings, Inc) or AliBaba as may be appropriate for national data residency requirements. Prior to processing data will only be accessible to the project technical staff using encrypted access.
LONG TERM ARCHIVING:
At the completion of the project the underlying data set of the raw color data and masked geographic location will be archived along with an associated .csv file containing a metadata reference. The intent is that the long term storage will be archived by the University of Sussex Library for the Sussex Humanities Lab, as well as mirror archive included as part of ancillary materials for a special section of the Leonardo Journal (published by MIT press) in Winter 2024. Leonardo currently hosts the data sets and other materials for those issues as part of their publication process. Finally as artist I will maintain a full archive of the project in perpetuity as part of my artistic records and papers.
CREATIVE COMMONS ARCHIVAL PUBLICATION OF DATA: The underlying dataset produced by the data collection project will be published under a CC BY-SA 4.0 (Creative Commons Attribution Share Alike or “copy-left” license), with all data contributors named as contributors unless they elect to be listed anonymously.
Other artists , participants and researchers may utilize the data set to create their own artworks or publications.
WORKS CREATED FROM THE DATA SET:
Any derivative artworks or publications created from the data set will be the IP of the creating artists to do with as they wish so long as the CC BY-SA terms for any distribution of the data set itself are honored.