Arduino with Python: Building a MicroPython Weather Node on the Nano ESP32

When makers search for Arduino with Python, they historically meant running C++ on the microcontroller while a host PC ran a PySerial script. In 2026, that paradigm has shifted. By leveraging the Arduino Nano ESP32 (ABX00092), you can now run MicroPython natively on official Arduino silicon, eliminating the need for a host PC while retaining the familiar Nano form factor. This guide walks through building a WiFi-connected environmental logger using MicroPython and an I2C BME280 sensor, complete with exact pin mappings, production-ready code, and bench-level debugging.

The Python-on-Arduino Landscape in 2026

Before wiring the breadboard, it is critical to understand which hardware and Python flavor you are actually deploying. Running Python on microcontrollers is not a monolith; the runtime, memory management, and hardware abstraction layers vary wildly between ecosystems.

Source: Arduino Nano ESP32 Cheat Sheet and MicroPython ESP32 Quick Reference.

Hardware BOM and Pin Mapping

The Arduino Nano ESP32 operates at 3.3V logic. This is a massive advantage for modern I2C sensors, as it eliminates the need for logic level shifters required by older 5V AVR boards.

Parts List

• Microcontroller: Arduino Nano ESP32 (Part: ABX00092) — ~$22.00
• Sensor: Adafruit BME280 I2C/SPI Breakout (Product ID: 2652) — ~$15.00
• Wiring: 22 AWG solid-core jumper wires (Red, Black, Blue, Yellow)
• Prototyping: Half-size solderless breadboard

Pin Mapping Table

The silkscreen on the Nano ESP32 uses traditional Arduino analog naming (A4/A5), but the underlying ESP32-S3 silicon uses specific GPIO numbers. MicroPython requires the GPIO numbers.

Flashing MicroPython and Project Setup

1. Download the Firmware: Navigate to the official MicroPython downloads page and grab the latest stable .bin release for the ESP32_GENERIC_S3 variant.
2. Erase and Flash: Open the Arduino IDE or use esptool.py to erase the flash. If using Thonny IDE (recommended), go to Tools > Options > Interpreter, select MicroPython (ESP32), and click 'Install or update MicroPython'.
3. Install the BME280 Library: MicroPython does not include sensor drivers in the base ROM. In the Thonny package manager (or via the REPL), install the bme280 package using the mip package manager: import mip; mip.install('bme280').
4. Verify I2C Address: The Adafruit BME280 (ID: 2652) defaults to I2C address 0x77. Generic clone boards often use 0x76. Check your board's silkscreen or run an I2C scan script before proceeding.

The Complete MicroPython Code

This script targets the Arduino Nano ESP32 (ABX00092). It initializes the I2C bus using the correct GPIO pins, connects to a 2.4GHz WiFi network, reads the BME280 sensor, and prints the telemetry. It includes robust try/except blocks to handle hardware and network faults without crashing the REPL.

import machine
import network
import time
import sys

# --- Hardware Configuration ---
# Nano ESP32 Silkscreen A4 is GPIO12, A5 is GPIO11
I2C_SDA_PIN = 12
I2C_SCL_PIN = 11
BME280_ADDR = 0x77  # Adafruit BME280 default

# --- Network Configuration ---
WIFI_SSID = 'Your_2.4GHz_SSID'
WIFI_PASS = 'Your_WiFi_Password'

# Initialize I2C Bus
try:
    i2c = machine.I2C(0, scl=machine.Pin(I2C_SCL_PIN), sda=machine.Pin(I2C_SDA_PIN), freq=400000)
    print(f'I2C Initialized. Devices found: {[hex(i) for i in i2c.scan()]}')
except Exception as e:
    print(f'FATAL: I2C Initialization failed: {e}')
    sys.exit(1)

# Import and Initialize BME280
try:
    import bme280
    sensor = bme280.BME280(i2c=i2c, address=BME280_ADDR)
    print('BME280 Sensor initialized successfully.')
except ImportError:
    print('FATAL: bme280 library not found. Run: import mip; mip.install("bme280")')
    sys.exit(1)
except OSError as e:
    print(f'FATAL: BME280 not responding at {hex(BME280_ADDR)}. Error: {e}')
    sys.exit(1)

# Connect to WiFi
wlan = network.WLAN(network.STA_IF)
wlan.active(True)

if not wlan.isconnected():
    print(f'Connecting to WiFi: {WIFI_SSID}...')
    wlan.connect(WIFI_SSID, WIFI_PASS)
    
    timeout = 15
    start_time = time.time()
    while not wlan.isconnected():
        if (time.time() - start_time) > timeout:
            print('FATAL: WiFi Connection Timed Out.')
            sys.exit(1)
        time.sleep(1)
        print('.', end='')

print(f'\nWiFi Connected! IP Address: {wlan.ifconfig()[0]}')

# Main Telemetry Loop
try:
    while True:
        temp_c, pressure_hpa, humidity_pct = sensor.read_compensated_data()
        print(f'Temp: {temp_c:.2f}C | Pressure: {pressure_hpa:.2f}hPa | Humidity: {humidity_pct:.2f}%')
        time.sleep(5)
except KeyboardInterrupt:
    print('\nLogging stopped by user.')
except Exception as e:
    print(f'Runtime Error in main loop: {e}')

Debugging: Exact Error Strings and Ranked Causes

Embedded Python strips away the luxury of desktop stack traces. When the Nano ESP32 throws an error at the hardware abstraction layer, you need to know exactly what the OS-level errno means. Here are the two most common failure modes for this specific build.

Error 1: OSError: [Errno 19] ENODEV

Context: This occurs during the bme280.BME280() initialization or the first I2C read.

Ranked Causes:

1. Wrong I2C Address: You are using a generic BME280 clone that defaults to 0x76 instead of the Adafruit 0x77. Fix: Change BME280_ADDR to 0x76.
2. Missing Pull-up Resistors: The internal ESP32-S3 pull-ups are weak (~45kΩ). If your breakout board lacks physical 4.7kΩ pull-ups on SDA/SCL, the bus will float. Fix: Add 4.7kΩ resistors from SDA and SCL to 3.3V.
3. Logic Level Damage: You accidentally wired the BME280 VIN to the Nano ESP32's 5V pin, frying the sensor's I2C transceiver. Fix: Replace sensor, verify 3.3V rail.

Error 2: OSError: [Errno 110] ETIMEDOUT

Context: This occurs during the wlan.connect() phase.

Ranked Causes:

1. 5GHz WiFi Network: The ESP32-S3 only supports 2.4GHz 802.11 b/g/n. It physically cannot see a 5GHz or 6GHz SSID. Fix: Ensure your router broadcasts a 2.4GHz band and use that SSID.
2. WPA3 Enterprise Security: MicroPython's native network stack struggles with WPA3-SAE or Enterprise (RADIUS) authentication out of the box. Fix: Use WPA2-PSK (Personal) for the IoT VLAN.
3. DHCP Exhaustion: Your router has run out of local IP leases. Fix: Reboot router or assign a static IP in the code.

Extending and Simplifying the Build

Once the baseline telemetry is printing to the REPL, you have two distinct paths for modifying the project based on your deployment environment.

How to Extend (Production IoT Deployment)

To make this a true IoT node, strip out the print() statements and integrate the umqtt.simple library. Configure the Nano ESP32 to publish a JSON payload to a local Mosquitto broker (e.g., running on a Raspberry Pi 5) every 60 seconds. You can also implement Deep Sleep using machine.deepsleep(60000) between reads. The ESP32-S3's RTC memory will retain the WiFi credentials, dropping the average current draw from ~80mA to under 15µA, allowing months of runtime on a single 18650 Li-ion cell.

How to Simplify (Educational/Bench Use)

If you are using this purely to learn Python syntax on hardware and do not care about environmental data, remove the BME280 entirely. The ESP32-S3 includes an internal temperature sensor. You can read it directly via the esp32 module without any external wiring:

import esp32
# Returns temperature in Fahrenheit, convert to Celsius
temp_c = (esp32.raw_temperature() - 32) * 5.0 / 9.0
print(f'Internal Core Temp: {temp_c:.1f}C')

This simplifies the hardware BOM to just the $22 Nano ESP32 board and a USB-C cable, making it the most cost-effective entry point for learning embedded Python in 2026. For deeper hardware schematics, always refer to the official Arduino documentation and the Adafruit BME280 learning guide.