Arduino+ESP32 Library Setup

We created an open-source client library to help you connect from embedded / IoT projects. Like this you can easily integrate AI training, inference and data flows into your applications.

Yes, you can do aicuflow data ingress from an ESP-32 Microcontroller!

Resources

IDE Setup Guide / Detail (this page)
Tutorial: Arduino Beginner Tutorial
Tutorial: Basic Sensor Plots on Arduino-ESP32 TFT eSPI Display
Tutorial: Training models from Arduino-ESP32 sensor data in aicuflow and visualizing it
More coming soon! Promise. :)

Our "embedded os" and library help you collect hardware sensor data easily.

It also adds cool features to ESP32 microcontrollers (see github).

Supported Board Specification

We tested with the ESP32-TTGO-T1 and T-Display-S3. Here's a pinout (graphic by otronic):

Other boards do work aswell, no worries! Check the github for details.

Installation

The installation is split in parts, of which you may skip some, depending on your current setup.

Arduino IDE with ESP32

Download and install the right version of Arduino IDE (Linux|Windows|MacOS)

On linux you might need to install fuse before, and add the user to dialout group (for Serial communication)
```
sudo apt install libfuse2t64
sudo usermod -aG dialout $USER
```
Possibly you have to reboot after this.

Open Arduino IDE.

On linux you might need to run (adjust your version)

chmod +x ./arduino-ide_2.3.7_Linux_64bit.AppImage
./arduino-ide_2.3.7_Linux_64bit.AppImage --no-sandbox

Select the board. Select "ESP 32 Dev Board" or a fitting model you are using.

In case that doesn't work, you can go to
File > Preferences > Additional Board Manager URLs (windows, linux)
or Arduino IDE > Settings > Additional Board Manager URLs (on mac)
and add:
```
https://raw.githubusercontent.com/espressif/arduino-esp32/gh-pages/package_esp32_index.json
```
Then you restart and go to Tools > Board manager, search for esp32 and add version 3.1.0 (tested; 3.3.5 might fail).
After that you can add a board like TTGO T1 if you have a LCD display.
On Windows, you might need to install drivers
For example CP210x Drivers for Windows

Installing the library

Clone the official repository for example sketches and the library header files:

git clone https://github.com/AICU-HEALTH/aicuflow-arduino

The library will soon also be installable from Arduino itself.

In upcoming versions, you will be able to train any ai model, run any flow with it.

For now, there's a tutorial:

Collecting, visualizing and training models from sensor data in aicuflow

Quickstart

Open the github repository for details.

The Aicuflow Lab

We bought a few dozen esp32s from a manufacturer at first, to experiment with them and let the team gather data within our platform.

ESP32s are "pretty cool" according to our team member Finn for a few reasons:

240MHz of performance at low energy consumption
WiFi and Bluetoot-LE integrated and straight forward
USB-C Port (the old ones didn't have it though)
Nice Analog capabilities (you can broadcast analog TV with it)

If we'd ever open an aicuflow merchandise shop, you could definitely buy these boards there too!

A few more ESP32 fun facts

You can use a very imprecise Hall Sensor to detect Magnetism on older devices
You can use the ESP32s Digital2Analog converter as a Software defined Radio for low frequencies
You can control an analog TV with a ESP32 pins
You can craft, view, transmit raw network packets for WIFI, BT and more

Many reasons to be excited about using this board with aicuflow!

Want to contribute?

Reach out to Finn for this.

Helpful links

Png to C++ Headerfile Converter - display images on TFT Screen
Png to Base64 Converter - for images in html on device
CSS Minifier - for hosting sites on device