Firmware for On-device ML on nRF54L15 MCU based BLE wearable device

Note: This firmware is compatible with the latest version of nRF connect SDK (3.1.1)

Building the device firmware (locally)

1. Install the nRF Connect SDK.

2. Clone this repository:

   $ git clone https://github.com/ramkumarkoppu/firmware-nordic-nrf54l15dk

3. Build the application:

   $ west build -b nrf54l15dk/nrf54l15/cpuapp --build-dir build_debug -- "-DCONF_FILE=prj_debug.conf" "-DSB_CONF_FILE=sysbuild_debug.conf"

Flashing (command line)

1. Connect the board and power on.

2. Flash the board controller firmware:

   $ west flash --build-dir build

Hardware Details

The Nordic Semiconductor nRF54L15 DK is a development board featuring the nRF54L15 SoC — part of the nRF54L Series — which offers excellent performance and ultra-low power consumption with Bluetooth Low Energy (BLE) connectivity. The DK allows for development of firmware for nRF54L15 SoC, and includes support for both baremetal and Zephyr RTOS based development.

This devkit does not include onboard sensors, so for Gesture classification experiments, connect the X-NUCLEO-IKS02A1 shield for accelerometer input. If you prefer a different sensor, you can modify this firmware (built with nRF Connect SDK and Zephyr RTOS) to support any Zephyr-compatible accelerometers.

With software in place, here's how to wire and connect the devkit.

Pictures of my Hardware Setup:

Wire Connections:

• Red wire: Power (1.8V)
• Black wire: Ground
• Green wire: I2C data
• Yellow wire: I2C clock
• White USB-C cable: Connection to the VS code