This repo holds files necessary to run PiLamps, a friendship lamp that acts as a way to stay connected with loved ones. The lamp is an IoT device, running on a Raspberry Pi connected to a touch sensor and LED strip, and powered by AWS IoT Core and SNS.

There are various folders and files in this repo: /ansible - WIP. For configuration of the RPi /app - The actual application Python files /config - WIP. For wifi configuration. /packer - Packer .hcl to create a Docker image used with testing and CI pipeline /terraform - Terraform files for creating IoT Core infrastructure

At this stage PiLamp is completely working, though I'd like to add a few new features (more on this below). My family and I use 3 lamps spread across the globe and have had no issues. With that said, there are some things to note:

1. Price: Using AWS will have some fees, though unless you have dozens of devices, the total monthly costs should only be a few dollars (<$5)

2. Physical Setup: There is a decent amount of manual setup required concerning the physical device (soldering and wiring skills. I've used RPi Zero through v4 devices to run PiLamps, along with ws2812x LEDs, a touch sensor, and shell to house everything. (This 3D printed lamp shell to be exact: https://www.thingiverse.com/thing:4129249)

3. Digital Setup: The RPi's OS and wifi will need to be configured manually. Once you can ssh into the RPi you can then pull this repo to your device.

How to setup

1. Once you're able to ssh into your RPi, clone this repo to a directory

2. In /terraform, edit the variables.tf file with your values, or use a .tfvars file.

3. Run 'setup.sh', which will do a few key things such as check for required packages (Python, AWSCLI, Terraform, pip packages, etc.), run terraform apply with auto-approve, creates certificates in ~/certs, and sets necessary environment variables.

    Note:
    LAMP_COLOR should be the name of a [CSS Matplotlib color](https://matplotlib.org/stable/gallery/color/named_colors.html#css-colors)


4. Start the program with a command similar to:
    sudo -E python3 ./Pi-Lamp/app/pi-lamp.py --endpoint \$LAMP_ENDPOINT --ca_file ~/certs/AmazonRootCA1.pem --cert ~/certs/cert.pem --key ~/certs/private.key --client_id \$LAMP_NAME --topic \$LAMP_TOPIC --count 0

    Note:
    Yes, this has to be run as sudo due to requirements from the rpi_ws281x package.

Future Aditions

1. More pipeline testing (docker)
2. Include dynamodb + lambda logging in Terraform
3. Add feature for when lamps touched at same time