Arduino Nicla Sense ME: Environmental Sensing at the Edge

The Arduino Nicla Sense ME is one of the smallest yet most sensor-packed boards in the Arduino ecosystem, purpose-built for environmental sensing at the edge. Measuring just 22.86 × 22.86 mm — roughly the size of a 50-paise coin — the Nicla Sense ME packs nine axes of motion data, a gas sensor, pressure/temperature/humidity readings, and an AI-ready microcontroller all on a single PCB. For Indian engineers and researchers working on smart agriculture, indoor air quality, wearable health monitors, or industrial condition monitoring, this board represents a significant leap in miniaturized sensing capability.

Table of Contents

• Hardware Overview: What’s Inside the Nicla Sense ME?
• Sensor Breakdown: Nine Axes + Gas + Climate
• BHI260AP: The Sensor Fusion Powerhouse
• Getting Started with Arduino IDE
• Environmental Sensing Projects
• Edge ML with the Nicla Sense ME
• Nicla Sense ME vs Other Sensing Platforms
• FAQ

Hardware Overview: What’s Inside the Nicla Sense ME?

The Nicla Sense ME is built around the Nordic nRF52832 SoC as the application processor (64 MHz Cortex-M4F, 512 KB flash, 64 KB RAM) and the Bosch BHI260AP as a dedicated AI-enabled motion and sensor co-processor. Here’s the full hardware summary:

• Application MCU: Nordic nRF52832 — 64 MHz ARM Cortex-M4F with FPU, Bluetooth 5.0 BLE
• Sensor Hub: Bosch BHI260AP — 32-bit Synopsys ARC EM4 @ 48 MHz, 50 KB RAM for ML models
• IMU: Bosch BMI270 — 6-axis (3-axis accelerometer + 3-axis gyroscope), ultra-low power, step/activity detection
• Magnetometer: Bosch BMM150 — 3-axis geomagnetic sensor (completes the 9-DOF IMU)
• Gas Sensor: Bosch BME688 — VOC, CO2-equivalent, temperature, humidity, barometric pressure
• Connectivity: BLE 5.0 via nRF52832
• Power: 1.8V I/O, onboard LDO, battery charging via ESLOV connector
• Form Factor: 22.86 × 22.86 mm, half-hole castellated edges for SMD mounting
• Interface: ESLOV connector (I2C) for daisy-chaining with other Nicla boards or Arduino hosts

The castellated edges allow the Nicla Sense ME to be soldered directly to a custom PCB as a module — making it ideal for product development and prototyping of miniaturized devices where a full development board footprint is unacceptable.

Sensor Breakdown: Nine Axes + Gas + Climate

Let’s go deeper into each sensor and what data it provides:

BMI270: 6-Axis IMU

The BMI270 is Bosch’s ultra-low-power IMU, optimized for wearable devices. Key specs:

• Accelerometer: ±2/±4/±8/±16 g range, 16-bit resolution
• Gyroscope: ±125/±250/±500/±1000/±2000 °/s range, 16-bit resolution
• Power: 685 µA (both active) — runs for months on a coin cell in low-duty-cycle mode
• Built-in step counter, activity recognition (stationary, walking, running, cycling)
• Significant motion detection, no-motion detection — wakes the system only when needed

BMM150: 3-Axis Magnetometer

Adds compass capability, completing the 9-DOF sensor suite. When fused with IMU data via the BHI260AP, you get absolute heading, full 3D orientation, and indoor navigation capability without GPS.

BME688: 4-in-1 Environmental Sensor

This is the star sensor for environmental monitoring:

• Temperature: ±1°C accuracy, -40 to +85°C range
• Humidity: ±3% RH, 0–100% range
• Pressure: ±0.6 hPa absolute accuracy (altitude resolution: ~0.5 m)
• Gas sensor: Metal Oxide Semiconductor (MOS) sensor detecting VOCs (Volatile Organic Compounds), ethanol, CO2-equivalent (eCO2)

Bosch’s BSEC (Bosch Sensortec Environmental Cluster) library processes raw BME688 data into interpreted outputs: air quality index (0–500), CO2 equivalent (ppm), breath VOC (ppm), and a stabilization indicator to show when the gas sensor has reached operating temperature for reliable readings.

BHI260AP: The Sensor Fusion Powerhouse

The BHI260AP is what separates the Nicla Sense ME from a simple breakout board collection. It’s a dedicated sensor fusion and AI co-processor with its own embedded Synopsys ARC EM4 processor running at 48 MHz and 50 KB of RAM for loading ML models.

Hardware Sensor Fusion

The BHI260AP runs Bosch’s proprietary 9-DOF fusion algorithm, producing:

• Game Rotation Vector: Quaternion (no magnetic north reference, drift-free for gaming/VR)
• Geomagnetic Rotation Vector: Quaternion referenced to magnetic north
• Gravity Vector: Current gravitational direction
• Linear Acceleration: Acceleration with gravity removed
• Euler angles: Roll, pitch, yaw in degrees

All these virtual sensors are computed on the BHI260AP, delivered to the nRF52832 via SPI — the application processor never needs to run fusion math, keeping power consumption extremely low.

AI on the BHI260AP

The BHI260AP can be programmed with custom ML models via Bosch’s BHY2 firmware loading mechanism. You can deploy Edge Impulse models (anomaly detection, gesture recognition) directly into the BHI260AP’s 50 KB RAM, running inference at ultra-low power even while the main nRF52832 sleeps.

Getting Started with Arduino IDE

1. Install Board Package: In Arduino IDE → Board Manager, search for “Arduino Mbed OS Nicla Boards” and install it.
2. Select Board: Tools → Board → Arduino Mbed OS Nicla Boards → Arduino Nicla Sense ME
3. Install Libraries: In Library Manager, install:
   • Arduino_BHY2 — the official sensor access library
   • Arduino_BHY2Host — for using Nicla as a shield on another Arduino board
4. Read All Sensors:

#include "Arduino_BHY2.h"

Sensor temperature(SENSOR_ID_TEMP);
Sensor humidity(SENSOR_ID_HUM);
Sensor pressure(SENSOR_ID_BARO);
Sensor gas(SENSOR_ID_GAS);
SensorXYZ accel(SENSOR_ID_ACC);
SensorXYZ gyro(SENSOR_ID_GYRO);
SensorQuaternion rotation(SENSOR_ID_RV);

void setup() {
  Serial.begin(115200);
  BHY2.begin();
  temperature.begin();
  humidity.begin();
  pressure.begin();
  gas.begin();
  accel.begin();
  gyro.begin();
  rotation.begin();
}

void loop() {
  BHY2.update();
  Serial.print("Temp: "); Serial.print(temperature.value()); Serial.println(" C");
  Serial.print("Humidity: "); Serial.print(humidity.value()); Serial.println(" %");
  Serial.print("Pressure: "); Serial.print(pressure.value()); Serial.println(" hPa");
  Serial.print("Gas: "); Serial.println(gas.value());
  delay(1000);
}

Environmental Sensing Projects

Indoor Air Quality Monitor

The BME688 with BSEC library outputs an IAQ (Indoor Air Quality) score from 0 (excellent) to 500 (extremely polluted). Build a BLE-connected AQI monitor that logs data to your phone via the Arduino IoT Cloud app. Add a small display (OLED or e-ink) on a custom PCB for a professional-looking product.

Smart Agriculture Sensor Node

Deploy multiple Nicla Sense ME nodes across a greenhouse. Each node measures temperature, humidity, and VOC levels (indicating plant stress from ethylene off-gassing or pesticide residue). Nodes transmit data via BLE to a central Arduino gateway connected to WiFi. This architecture is used in precision agriculture installations.

Wearable Activity Tracker

The BMI270’s built-in step counter and activity recognition, combined with the BHI260AP’s low-power fusion, make the Nicla Sense ME ideal for wearable fitness trackers. The 9-DOF output enables swim stroke detection, fall detection for elderly care, and rehabilitation exercise monitoring — all on a board small enough to embed in a wristband.

Industrial Vibration and Condition Monitoring

Mount the Nicla Sense ME on a motor housing. The BHI260AP streams accelerometer data to an Edge Impulse anomaly detection model. When bearing wear begins to produce characteristic vibration frequencies, the model triggers an alert via BLE before catastrophic failure occurs — predictive maintenance in a 23mm package.

Edge ML with the Nicla Sense ME

The Nicla Sense ME integrates seamlessly with Edge Impulse for deploying custom ML models. Here’s the end-to-end workflow:

1. Data Collection: Use the Edge Impulse CLI with the arduino-nicla-sense-me target to stream sensor data directly to Edge Impulse Studio for labelling.
2. Feature Engineering: Edge Impulse auto-selects appropriate signal processing blocks (MFCC for audio, Flatten/Spectral Analysis for IMU, Raw for environmental sensors).
3. Model Training: Train a classifier or anomaly detector in the browser. Edge Impulse optimizes the model for the BHI260AP’s 50 KB constraint.
4. Deployment: Export as an Arduino library. Include it in your sketch — inference runs on the BHI260AP co-processor, leaving the nRF52832 free for BLE communication.
5. Continuous Learning: With Edge Impulse’s deployment API, you can push updated models wirelessly via BLE OTA to deployed Nicla nodes in the field.

Common ML use cases validated on the Nicla Sense ME: gesture recognition (hand gestures for UI control), fall detection (gyroscope + accelerometer fusion), air quality anomaly detection (BME688 pattern analysis), and predictive maintenance (motor vibration classification).

Nicla Sense ME vs Other Sensing Platforms

FAQ

Can the Nicla Sense ME be used as a shield on another Arduino?

Yes! The Nicla Sense ME can operate as a BLE or I2C peripheral to a host Arduino board. Using the Arduino_BHY2Host library and the ESLOV cable (I2C), any Arduino with I2C can read all sensor data from the Nicla — effectively making it a smart sensor module for your existing project.

How long does the Nicla Sense ME run on a battery?

In low-power mode with BLE advertising every second and sensors sampling at 1 Hz, the Nicla Sense ME draws approximately 2–5 mA. With a 200 mAh LiPo battery, you can expect 40–100 hours of operation. With the main MCU in deep sleep and only the BHI260AP active for motion detection, current drops below 100 µA, extending battery life to months.

How do I update the BHI260AP firmware?

The BHI260AP firmware is loaded at boot from the nRF52832’s flash storage. When you use the Arduino_BHY2 library, it automatically uploads the latest BHY2 firmware to the BHI260AP on startup if a newer version is detected. You don’t need special tools — just updating the library version handles firmware updates transparently.

Is the Nicla Sense ME waterproof?

The bare board is not waterproof. However, the BME688 sensor requires contact with ambient air for gas sensing, so full encapsulation is counterproductive. For outdoor or high-humidity deployments, use a conformal coating on the PCB while leaving the BME688 sensor opening uncoated, and place the assembly in a housing with a Gore-Tex membrane vent.

Can I do BSEC air quality processing on Nicla Sense ME?

Yes, Bosch’s BSEC library is available for the nRF52832 (Arduino Mbed OS). It runs on the main MCU and processes raw BME688 data into IAQ scores, eCO2, and breath VOC readings. Note that BSEC requires a 5-minute warm-up period for the gas sensor heater plate to stabilize before accurate readings are produced.

Ready to build your environmental sensing project? Explore our full range of Arduino boards and sensors at Zbotic — from the Nicla Sense ME to affordable BME280 and DHT20 modules. Fast delivery across India for all your IoT and maker needs.