Add AuraSense application

application/AuraSense/.gitattributes

@@ -0,0 +1,11 @@
+* text=auto
+
+*.bat text eol=crlf
+*.ps1 text eol=crlf
+*.s37 binary
+*.tflite binary
+*.png binary
+*.jpg binary
+*.jpeg binary
+*.so binary
+*.p12 binary

application/AuraSense/.gitignore

@@ -0,0 +1,51 @@
+# OS / editor
+.DS_Store
+Thumbs.db
+.idea/
+.vscode/
+*.swp
+*.tmp
+
+# Python
+__pycache__/
+*.pyc
+.venv/
+venv/
+
+# Android / Gradle
+.gradle/
+.kotlin/
+build/
+**/build/
+local.properties
+*.iml
+
+# Simplicity Studio / generated build output
+**/.pdm/
+**/.settings/
+**/.uceditor/
+**/cmake_gcc/
+**/GNU ARM v12.2.1 - Default/
+**/build.ninja
+**/.ninja_deps
+**/.ninja_log
+**/CMakeFiles/
+**/CMakeCache.txt
+
+# Backup / scratch files
+*.bak
+*.bak0
+*.orig
+
+# Archives / binaries produced during local work
+*.zip
+*.bin
+*.hex
+*.out
+*.axf
+*.map
+*.apk
+
+# Node
+node_modules/
+**/node_modules/

application/AuraSense/README.md

@@ -0,0 +1,117 @@
+# AuraSense EIC 2026
+
+AuraSense is a smart baby-monitoring project built for the EIC 2026 context. It combines an EFR32xG26 firmware application, a custom Android caregiver app, and an edge AI audio model so a nursery device can listen for baby distress sounds and notify a caregiver over Bluetooth Low Energy (BLE).
+
+## What This Project Does
+
+In simple terms:
+
+1. The EFR32xG26 board listens to nearby baby audio.
+2. The firmware runs an on-device TinyML model.
+3. The board classifies the sound as background, cry, laugh, or sad/distress.
+4. The result is sent to the Android app over BLE.
+5. The app shows live status, confidence, room temperature, and alerts the caregiver when distress is detected.
+
+## Who This Repo Is For
+
+- Students or judges who want to understand the project quickly
+- Developers who want to build the firmware or app
+- Anyone working with Silicon Labs EFR32xG26 hardware
+
+## Hardware Target
+
+This firmware project is configured for:
+
+- Silicon Labs EFR32xG26 family
+- Device: `EFR32MG26B510F3200IM68`
+- Board component in project metadata: `brd2608a`
+
+If you use a different EFR32xG26 board, you may need to adjust microphone, sensor, or pin configuration before building.
+
+## Repository Layout
+
+```text
+AuraSense_EIC_2026/
+  firmware/
+    aura-baby-monitor-soc/     # Main EFR32xG26 firmware project
+    releases/                  # Ready-to-flash .s37 firmware image
+  mobile/
+    aurasense-android-app/     # Android caregiver application
+  model/
+    training/                  # Training script and training notes
+    artifacts/                 # Deployable TFLite model
+    datasets/                  # Dataset pointer / dataset notes
+```
+
+## Quick Start
+
+If you only want to try the project as fast as possible:
+
+1. Flash `firmware/releases/aura_baby_monitor.s37` to your EFR32xG26 board.
+2. Open `mobile/aurasense-android-app` in Android Studio.
+3. Build and install the Android app on an Android phone.
+4. Power the board, pair/connect over BLE, and open the Baby Cry Monitor screen.
+
+## Build Requirements
+
+### Firmware
+
+- Simplicity Studio 6
+- Silicon Labs Simplicity SDK `2025.12.1`
+- AI/ML extension `2.2.0`
+- Supported EFR32xG26 hardware
+- J-Link / Simplicity Commander for flashing
+
+### Android App
+
+- Android Studio
+- JDK 17
+- Android SDK for `compileSdk 36`
+- Android device with BLE support
+
+## Main Files You Will Care About
+
+### Firmware
+
+- `firmware/aura-baby-monitor-soc/ml_ble_classifier.slcp`
+- `firmware/aura-baby-monitor-soc/app.c`
+- `firmware/aura-baby-monitor-soc/audio_classifier.cc`
+- `firmware/aura-baby-monitor-soc/config/tflite/baby_cry_int8_DEPLOY.tflite`
+
+### Android App
+
+- `mobile/aurasense-android-app/mobile/build.gradle.kts`
+- `mobile/aurasense-android-app/mobile/src/main/java/com/siliconlabs/bledemo/features/demo/babycry/BabyCryMonitorActivity.kt`
+- `mobile/aurasense-android-app/mobile/src/main/res/values/strings_baby_cry.xml`
+
+### Model
+
+- `model/training/training.py`
+- `model/artifacts/baby_cry_int8_DEPLOY.tflite`
+- `model/datasets/datasets.txt`
+
+## How The System Works
+
+- The firmware performs audio feature extraction and TensorFlow Lite Micro inference on the device.
+- BLE notifications carry the inference result to the phone.
+- The Android app turns those predictions into a caregiver-friendly monitoring screen and alert flow.
+- Temperature data is also surfaced when available through the firmware configuration.
+
+## Notes For Public Use
+
+- This repo intentionally excludes local IDE files, build output, and generated machine-specific folders.
+- The Android app is a customized fork of Silicon Labs' mobile app shell, adapted here for AuraSense monitoring.
+
+## Folder Guides
+
+Each main folder contains its own `README.md` with detailed instructions:
+
+- `firmware/README.md`
+- `firmware/aura-baby-monitor-soc/README.md`
+- `firmware/releases/README.md`
+- `mobile/README.md`
+- `mobile/aurasense-android-app/README.md`
+- `model/README.md`
+- `model/training/README.md`
+- `model/artifacts/README.md`
+- `model/datasets/README.md`

application/AuraSense/firmware/README.md

@@ -0,0 +1,27 @@
+# Firmware
+
+This folder contains everything needed on the embedded side of AuraSense.
+
+## What Is Here
+
+- `aura-baby-monitor-soc/` - the main Simplicity Studio firmware project
+- `releases/` - prebuilt firmware image for fast flashing
+
+## What The Firmware Does
+
+- Records audio from the on-board microphone path
+- Runs TensorFlow Lite Micro inference on the EFR32xG26
+- Sends classification results over BLE
+- Exposes basic control commands for the phone app
+- Can publish room-temperature data when the board configuration supports it
+
+## If You Are New
+
+Start here:
+
+1. Read `aura-baby-monitor-soc/README.md`
+2. If you want the fastest demo path, use `releases/README.md`
+
+## Hardware Focus
+
+The included project metadata targets the EFR32xG26 family and is configured around the Silicon Labs board setup referenced in the `.slcp` project file.