The coffee situation bot observes the current office coffee machines situation as it is important to know the status of the coffee situation
On a slack channel, the bot app listens for specific keywords eg. coffee, kahavia.. and on a match it responds with a situation message that has an image link attached to it. The image is taken by the coffee observation device and it is from the office coffee machine.
The user sees couple of things
The slack bot responses to different events
| Event | Message type | Response |
|---|---|---|
| Coffee keyword | On a channel |  |
| Joke | Direct message |  |
| Status | Direct message |  |
| Connection error | - |  |
| Unspecified error | - |  |
The responses have some randomness on the icon and the affix text
The observation device takes an image and reports it as a temporary link
Example situation image with some visual blocking
- Image contents vary depending on the observation device positioning
- should the device positioning resulted images reveal too much of for example some corporate secrets or public spaces the resulting image can be filtered with visual blocks or even face blurring ✨
- images have a limited lifetime
✨ results may vary
- Help: Prints this usage text
- Status: Checks if the coffee situation monitoring device is online
- Joke: Well, why not?
- List: Prints accepted coffee related keywords
Coffee keyword: Takes a photo of the current coffee situation
TODO: Description of the different device mode configurations
- In coming versions there should be some real coffee situation senseness
- automaticish area visual blocking
- improvements on the face blurring
- OpenCV related device stability
The app consists of a Rasbperry Pi device taking pictures and a Slack app reporting the results. Thus there are two different setups and configurations.
The slackbot app runs and installs with:
- python3
- cron
- pip3
- git
- Clone the repository
git clone https://github.com/lsipii/coffeesituation.git
- Install the bot requirements
For the slackbot hosting server the reqs are listed in the slackbot_requirements.txt file
pip3 install -r coffeesituation/slackbot_requirements.txt
- Copy the example configurations:
cp config/slackbotApp.example.json config/slackbotApp.json
- Configure at the very least
- SLACK_BOT_TOKEN: The slack access key, by which the bot connects to the Slack workplace
- SLACK_BOT_MAINTAINER: Slack user ID, eg. W012A3CDE
- COFFEE_BOT_TOKEN: The access key to the coffee observation device, self generated shared secret
- COFFEE_BOT_URL: The public access url of the coffee bot device, eg. https://xxXXxxXXXXxx.dataplicity.io
- Cronjob configurations
by
crontab -e
add lines, for example (fix the paths):
@reboot /path/to/coffeesituation/shell/ensureSlackBotAppRunning.sh
16 2 * * * /path/to/coffeesituation/shell/selfUpdate.sh
10,27 * * * * /path/to/coffeesituation/shell/ensureSlackBotAppRunning.sh
TODO: Raspberry Pi setups
The dependency installations are better described in the Dockerfile
Thought, multiliner and multistepper without OpenCV support would be
- Base requirements
sudo apt-get update -y && sudo apt-get install -y git python3-pip nginx motion git
- Motion configurations
sudo sed -i 's/daemon off/daemon on/g' /etc/motion/motion.conf && \
sudo sed -i 's/stream_localhost on/stream_localhost off/g' /etc/motion/motion.conf && \
sudo sed -i 's/stream_port 0/stream_port 8081/g' /etc/motion/motion.conf && \
sudo sed -i 's/output_pictures on/output_pictures off/g' /etc/motion/motion.conf && \
sudo sed -i 's/ffmpeg_output_movies on/ffmpeg_output_movies off/g' /etc/motion/motion.conf && \
sudo sed -i 's/stream_maxrate 1/stream_maxrate 25/g' /etc/motion/motion.conf && \
sudo sed -i 's/framerate 100/framerate 25/g' /etc/motion/motion.conf && \
sudo sed -i 's/width 352/width 640/g' /etc/motion/motion.conf && \
sudo sed -i 's/height 288/height 480/g' /etc/motion/motion.conf && \
sudo sed -i 's/start_motion_daemon=no/start_motion_daemon=yes/g' /etc/default/motion && \
sudo systemctl disable motion
- Clone the repository if not yet done
git clone https://github.com/lsipii/coffeesituation.git
- nginx configurations
sudo cp ./coffeesituation/docker/nginx/sites/default.conf /etc/nginx/sites-available/default && \
sudo mkdir -p /var/www/html/images && sudo cp ./coffeesituation/docker/nginx/404/404.jpg /var/www/html/images
- The deviceApp python requirements
pip3 install -r ./coffeesituation/requirements.txt
TODO: better decribe the setup using dataplicity,
- but for now it's pretty straightforward: https://www.dataplicity.com/devices/
- enable the wormhole
- get the public address and reference where called for a COFFEE_BOT_URL setting
TODO: setup using resin.io
- ...
- Copy the example device configuration file
cp ./coffeesituation/config/deviceApp.example.json ./coffeesituation/config/deviceApp.json
- Fill important parts of the json file
- app.host: the devices public address eg. the COFFEE_BOT_URL
- app.storage_driver: "local" or "S3"
- apiAccess.Coffee Related Communication And Relations Facilitator.api_token: the SlackBot access token, the COFFEE_BOT_TOKEN setting of the SlackBot
- storage.local or storage.S3
- Test if the app starts running
./coffeesituation/deviceApp.py
- Stop by ctrl-c
- Cronjob configurations
by a crontab editing command:
crontab -e
add lines:
@reboot /home/pi/coffeesituation/shell/ensureDeviceAppRunning.sh
15 2 * * * /home/pi/coffeesituation/shell/selfUpdate.sh
20,57 * * * * /home/pi/coffeesituation/shell/ensureDeviceAppRunning.sh
10,35,44 * * * * /home/pi/coffeesituation/shell/ensureRaspberryPiNetwork.sh
If using S3 storage, add also a:
2 * * * * /home/pi/coffeesituation/S3BucketGC_cronjob.py
- Finalize by a reboot
sudo reboot
Turtle photo: https://www.pexels.com/fi-fi/kuva/elain-elainkuvaus-kilpikonna-lahikuva-2289601/