Note

I am still in the process to getting everything working, so this project is not yet complete

About this project

Why

The reason for this project is two-fold. First and main reason is, is that I found that mold was forming in the corners of my bedroom in a new flat that I am renting and I wanted to find the cause. To do this I would have to track the room temperature and humididy. After analysing the data I could determine is I would have heat more, air the room more or if there moisture coming thought the wall.

Second is that I love programming, learing new tech and automation so I thought this would be a great change to combined them all together to produce a usefull soultion.

Prereqs

To get things running you need Java, Maven and Docker. Which version ??? this topic always sucks. I have gone through so many tutoial and they never work because my version is newer than the versions used when the tutioial was written. I don't really have a solution other that you should know a bit of:

• Java
• Maven
• Raspberry pi
• Databases
• Nginx / PHP

and later:

• Go
• Spring boot

Desription

So, the idea is that in the end I would have a Raspberry Pi that is running a db, some sort of service that accepts connections form a sensor and adds the values to the db. Then there should be some sort of web front end to view the temperature and humidity.

For the db I choose Postgresql, why? does is really matter? Whether you used H2, SQL lite, Mysql, Maria, Postgres for such simple project you can take your pick. I also plan to use Hiberate so then the db is really not an issue.

Next lets look at a sensor. I started with the wemos d1 mini the a DHT22 temperature-humidity sensor. Then to make things I bit more interesting and easier I moved to the Sonnoff Basic. Theo Arends has a great project that makes using the sensor just a matter of configureation. Plus with the sonoff you can do other things like switch a light off and on. Another added extra is the MQTT support and installing a broker on the Pi is also a no brainer.

Here is a link to his project: https://github.com/arendst/Sonoff-Tasmota

Then I wrote a very simple java service that listens for messages published from the sonoff and inserts the data in to the database.

For the frontend to make this quick and easy I used Highcharts js libraies to visualize the data. Adding nginx and PHP-5 to the Pi again no problem. All you have to do is write a simple php and html script to link it all together and now I can easily monitor the temperature and humidity in every room of my flat.

Here is a link to their stuff: https://www.highcharts.com/stock/demo

To develope the system what better way than to use Docker.

How to uses this project

• Check it out
• Build the service using maven
• Run docker compose
• either pre populate the db with come code or run some dummy sensors for a real-time effect Then when it all works:
• deploy every thing on a Pi and Sonoff(s)

Details at soon to come...

Next steps

The next steps are to use:

• Use Go for the sevice instead of java.
• Use Spring Boot, GraphQL and Highcharts to access and visualize the data.
• Try and deploy the project on Kubernetes (I already have a cluster of Pis running ;-)