NDP.onMatch() does not work with the accelerometer sensor

Official Hardware Nicla Family Nicla Voice
1

Hello, I have made a model that allows detecting falls using the accelerometer values ​​as data and uses the Syntiant processing block. The model was made in edge impulse.

image
image990×563 32.3 KB

I made a sketch in Arduino Ide to be able to obtain the result of the inference. The code shows through the serial monitor the label that was inferred

#include "NDP.h"
#include "Nicla_System.h"


const bool lowestPower = false;

void ledBlueOn(char* label) {
  if (!lowestPower) {
      Serial.println("Label received");
      nicla::leds.begin();
      nicla::leds.setColor(blue);
      delay(200);
      nicla::leds.setColor(off);
      Serial.println(label);
      nicla::leds.end();
  }
}

void setup() {
  Serial.begin(115200);
  nicla::begin();
  nicla::disableLDO();
  nicla::leds.begin();

  NDP.onError([]() {});
  NDP.onMatch(ledBlueOn);
  NDP.onEvent([]() {});
  Serial.println("Loading synpackages");
  Serial.println("This is a custom flash for sacha model");
  NDP.begin("mcu_fw_120_v91.synpkg");
  NDP.load("dsp_firmware_v91.synpkg");
  NDP.load("ei_model.synpkg");
  Serial.println("Packages loaded");
  NDP.getInfo();
  Serial.println("Configure mic");
  NDP.turnOnMicrophone();
  NDP.interrupts();

  nicla::leds.end();

  if (lowestPower) {
    NRF_UART0->ENABLE = 0;
  }
}

void loop() {
}

But no results appear on the serial monitor.

Packages loaded
dsp firmware version: 19.2.0
package version: Model generated by Edge Impulse
pbi version: 3.2.3-0
num of labels: 2
labels: NN0:ADL, NN0:FALL
total deployed neural networks: 1
Configure mic

I made another model to see if it was the board that was not working correctly, this model detected a predefined word through the microphone. In this model it did show me the inference on the serial monitor

Packages loaded
dsp firmware version: 19.2.0
package version: Model generated by Edge Impulse
pbi version: 3.2.3-0
num of labels: 4
labels: NN0:no_sacha, NN0:noise, NN0:si_sacha, NN0:unknown
total deployed neural networks: 1
Configure mic
Label received
NN0:unknown
Label received
NN0:noise
Label received
NN0:noise

I would like some help on how I can make the model used by the IMU sensor show the inference through the serial monitor, as I later want to send that value via Bluetooth to an App. Thank you.

2

Sorry I cannot help witht the IMU, but for sending the results a simple method is to just update the advertised BLE name with the result appended.
See the example ei_model.ino code below

For a reciever I use an ESP32C3 see
https://www.forward.com.au/pfod/HomeAutomation/Wifi2BLE/ESP32C3_BLE_WIFI/index.html

Writing an app to do something with the results is non-trivial.
To just see the results on your phone is to use the ESP32C3 to bridge between the BLE advert and a webpage that you can open on your mobile phone.

For app building you could try
https://appinventor.mit.edu/
but I have not used it myself.

ei_model.ino (5.4 KB)