MycoRover

Devlog #2
From an idea to a scientific foundation
After publishing my first devlog, I realized something important.
Before building the rover or writing any AI, I first needed to answer a much bigger question:
Which microorganisms actually improve soil health, and under what conditions do they thrive?
Instead of immediately ordering components or designing hardware, I decided to spend this week building the scientific foundation of MycoRover.
I started creating a microorganism database that will eventually become the knowledge base for the AI recommendation system. Rather than trying to include every beneficial microorganism, I chose to focus on a carefully selected set that are well studied and commonly used in sustainable agriculture.
For each microorganism, I’m documenting:

Its category and primary function

Which nutrient it helps improve (Nitrogen, Phosphorus, or Potassium)

Suitable crops

Preferred soil pH

Moisture requirements

Temperature range

Electrical conductivity (EC) preferences

Whether the rover can realistically create or maintain those conditions

Scientific references supporting the information

So far, I’ve researched and documented the first five microorganisms. While doing this, I also realized that the rover shouldn’t “add nutrients” directly. Instead, its main responsibility should be to create favorable environmental conditions that allow beneficial microorganisms to establish themselves and naturally improve soil health.
This research phase has already changed how I think about the project. Originally, I imagined the AI would simply predict which microorganism to use. Now I’m designing it as a recommendation system built on scientific evidence, making every recommendation explainable and backed by research.
I also spent around 1 hour or so ,researching soil microbiology, reading scientific literature, and organizing the database that will later power the AI.
The next milestone is to complete the microorganism database, validate each entry using reliable scientific sources, and begin designing the recommendation engine that will connect the soil testing machine to the autonomous rover.

DEVLOG #1

For the past few days, I’ve been thinking about one question:
Can we reduce the use of chemical fertilizers by helping beneficial microorganisms grow instead?
Most smart farming projects focus on detecting soil conditions and recommending fertilizers. I wanted to explore a different approach. Instead of feeding the soil with more chemicals, what if we could create the right conditions for naturally beneficial microorganisms to thrive?
That idea became the starting point for my project.
Right now, I’m calling it MycoRover.
The project will have three main parts:

A soil testing machine that analyzes soil samples by measuring pH, NPK, electrical conductivity (EC), moisture, and temperature.

A central AI hub running on a Raspberry Pi 5 that processes the collected data and recommends suitable beneficial microorganisms based on the soil conditions.

An autonomous rover that follows predefined paths between crop rows and carries out the treatment plan by creating favorable environmental conditions for those microorganisms.

At this stage, I haven’t started building any hardware yet. Today was all about planning the system before buying components or writing code.
I spent most of the day filling pages of my notebook with ideas.
First, I thought about how to collect a soil sample that actually represents the entire field instead of relying on a single random spot. I sketched a method of dividing the field into sections and combining samples to get a more representative average.

Next, I designed a rough concept for the soil testing machine. The idea is to place the mixed soil inside a testing chamber where multiple sensors can automatically measure pH, NPK, EC, moisture, and temperature before sending the readings to the AI hub.

Finally, I outlined the overall workflow of the project—from soil testing to AI analysis and finally to the rover carrying out the treatment plan.

This is still the very beginning, and I expect many parts of the design to change as I learn more about soil microbiology and robotics. My goal for the next few days is to research beneficial microorganisms, validate the scientific assumptions behind the project, and begin designing the first prototype.
I’m excited to see how far this idea can go.