Automated Gardening

Prior Gardening Projects and Forecasting

Last year I created an Arduino Mini Pro based plant watering control system.

The Garduino was designed to control one flowerpot. It indicates the room temperature, light intensity and soil moisture.
The watering system was designed as a one point moisture controller. Until a prevously set soil moisture value is reached, the watering system starts for a certain time.
This system imitates real environmental behavour as sequential soil aridification prevents the plant to rot at countinous high moisture level.

This year this system will be expanded to control an even bigger outside patch!

Patch Preperation

We started on preparing the patch in an allotment garden in Hamburg.
For this, we removed the turf in a 2 x 3 m area.
Some of the turf was placed on sparse sections, some ended up side down on the compost.
After that, eding stones were placed all around the patch to reach a certain sovereignty.

The patch was dug over and mixed up with finished compost to deliver needed organic and inorganic substrates.
Afterwards, we used some normal flower soil for the upper layer.
We did some watering and let the microorganisms and worms do the microcosmos work done.

Seeding & Preculturing

For better seeding and easy seperation of plants in their early stage, I printed myself a plant pricking tool (dibber) which you can find at

The seeds were orderd online at Bingenheimer Saatgut.
I ordered the following seeds:

Black Cherry (Cherry Tomato), Zuckertraube (Snack Tomato), Hellfruch (Tomato), Schönbrunner Gold (Physalis),
Yolo Wonder (Paprika), De Cayenne (Chili), Bijella (Salad), Cerbiatta (Salad)

The plants were bred in small seeding pots with just normal potting soil. They germed after around 1,5 weeks.

A LED panel makes sure the plants receive enough sun on rainy and cloudy days, which I bough on Amazon.


It’s a process, not a one-time-installation!

I am currently using the old Wemos D1 Mini, one among many IOT microcontrollers with integrated WiFi.

It receives data by a DHT22 (temperature and humidity) and a BH1750 (light intensity) sensor, which are recorded by different plattforms.
The data are logged by Blynk for smartphone control, and ThingSpeak to perform Matlab based analytics like these realtime plots:

Temperature & Humidity

Light intensity

Weather Forcasting