coffeebuddy

Do you use a paper based tally sheet to count your team's coffee consumption? Throw it away and use coffeebuddy!

Setup

Coffebuddy uses pdm to manage its python dependencies.

0. Install pdm

   curl -sSL https://pdm-project.org/install-pdm.py | python3 -

1. Install dependencies

Server

Hardware (recommendation)

• RaspberryPi 4
• External SSD to store database

Software

The server runs postgresql and uses certificated based authentication on client side. Details are described in Setting up postgresql database.

Client

Hardware (recommendation)

• RaspberryPi 3
• 7" touchscreen with 1024x600 pixel resolution
• Identiv uTrust 1401f PCSC cardreader
• 3D printed housing (see doc/housing.md) including:
  • a PIR for presence detection (see coffeebuddy/pir.py)
  • a multicolor LED for status indication (see coffeebuddy/illumination.py)

Software

1. Install cardreader dependencies If using a PCSC cardreader and a Debian based distributions:

   sudo apt install swig libpcsclite-dev pcscd pcsc-tools

2. Install Python dependencies

   pip install --user pdm
   
   pdm sync

3. Start production environment

   ./bin/run.py

Debugging

The development environment can be run using

FLASK_DEBUG=1 ./bin/run.py

Tests

Run tests with pytest -v tests

Application

The final application uses a Raspberry Pi attached to a 7" touchscreen. Thus, the HTML an CSS is optimized for a display with a resolution of 1024x600.

How to setup a client Raspberry Pi

1. Flash latest RapberryPi OS

2. Fix screen resolution:

   # /boot/firmware/config.txt
   [all]
   max_usb_current=1
   
   hdmi_force_hotplug=1
   hdmi_group=2
   hdmi_mode=87
   hdmi_cvt 1024 600 60 3 0 0 1
   hdmi_drive=2
   
   start_x=1
   

   And change dtoverlay=vc4-kms-v3d to

   dtoverlay=vc4-fkms-v3d
   

3. If display has to be rotated by 180° adjust /etc/X11/xorg.conf.d/40-libinput.conf

   Section "InputClass"
       Identifier "libinput touchscreen catchall"
       MatchIsTouchscreen "on"
       MatchDevicePath "/dev/input/event*"
       Driver "libinput"
       Option "CalibrationMatrix" "-1 0 1 0 -1 1 0 0 1"
   EndSection
   

   Rotate the display by using an autostart file

   # ~/.config/autostart/rotate.desktop
   
   [Desktop Entry]
   Name=Rotate
   Type=Application
   Exec=xrandr --output HDMI-1 --rotate inverted
   

4. Install required or helpful packages:

   sudo apt install swig libpcsclite-dev pcscd pcsc-tools  vim screen unclutter

5. Start chrome once and disable translation option.

6. Clone this repo and install requirements:

   git clone https://github.com/hackenbergstefan/coffeebuddy.git
   cd coffeebuddy
   curl -sSL https://pdm-project.org/install-pdm.py | python3 -
   pdm sync

7. Create and copy certificates for database connection:

   scp coffeebuddyserver:~/coffeebuddy/database/certs/ca/root.crt coffeebuddy01:~/.postgresql/
   scp coffeebuddyserver:"~/coffeebuddy/database/certs/client_coffeebuddy01/postgresql.*" coffeebuddy01:~/.postgresql/

8. Create and adjust config_<client>.py.

9. Enable autostart:

   mkdir ~/.config/autostart
   cp raspi/coffeebuddy.desktop ~/.config/autostart/

10. Enable gpio service:

    sudo systemctl enable pigpiod

Card reader

Coffeebuddy works with PCSC reader and with SPI RFID module "RC522". Latter is supported on Raspi by several python modules. Although mrfc522 is widely used it leads to a high CPU consumption when polling for card. pi-rc522 uses interrupt based SPI communication.

Both modules can be used and selected in config.py.