Arduino Tiny Machine Learning Kit

Arduino Tiny Machine Learning Kit merupakan kit pembelajaran Machine Learning (TinyML) berbasis Arduino Nano 33 BLE Sense. Dengan Board Arduino Nano 33 BLE Sense yang sudah terintegrasi dengan berbagai sensor sehingga kita dengan dapat membuat aplikasi seperti mendeteksi suara, gerakan, rotasi, tekanan barometrik, akselerasi, proximity, warna, dan intensitas cahaya. Kit ini juga sudah termasuk modul kamera OV7675 (dapat digunakan untuk pengenalan wajah) serta shield Arduino khusus yang dapat membantu untuk menciptakan proyek Machine Learning praktis dengan algoritma klasik dan neural network menggunakan TensorFlow Lite Micro.

Spesifikasi Arduino Nano 33 BLE Sense :

• Mikrokontroller : nRF52840 (datasheet)
• Tegangan operasi : 3.3V DC
• Tegangan Input : 21V DC (Maks.)
• DC Current per I/O Pin : 15 mA
• Clock Speed : 64MHz
• CPU Flash Memory : 1MB (nRF52840)
• SRAM : 256KB (nRF52840)
• EEPROM : No
• Digital Input / Output Pins : 14
• PWM Pins : all digital pins
• UART : 1
• SPI : 1
• I2C : 1
• Analog Input Pins : 8 (ADC 12 bit 200 k samples)
• Analog Output Pins : Only through PWM (no DAC)
• External Interrupts : all digital pins
• LED_BUILTIN : Pin 13
• USB : Native in the nRF52840 Processor
• IMU : BMI270 6-axis IMU Accelerometer and Gyroscope (datasheet)
• 16-bit
• 3-axis accelerometer with ±2g/±4g/±8g/±16g range
• 3-axis gyroscope with ±125dps/±250dps/±500dps/±1000dps/±2000dps range
• IMU : BMM150 3-axis Magnetometer (datasheet)
• 3-axis digital geomagnetic sensor
• 0.3μT resolution
• ±1300μT (x,y-axis), ±2500μT (z-axis)
• Microphone : MP34DT06JTR (datasheet)
• AOP = 122.5 dbSPL
• 64 dB signal-to-noise ratio
• Omnidirectional sensitivity
• –26 dBFS ± 3 dB sensitivity
• Gesture, light, proximity, color : APDS9960 (datasheet)
• Ambient Light and RGB Color Sensing with UV and IR blocking filters
• Very high sensitivity – Ideally suited for operation behind dark glass
• Proximity Sensing with Ambient light rejection
• Complex Gesture Sensing
• Barometric pressure : LPS22HB (datasheet)
• 260 to 1260 hPa absolute pressure range with 24 bit precision
• High overpressure capability: 20x full-scale
• Embedded temperature compensation
• 16-bit temperature data output
• 1 Hz to 75 Hz output data rateInterrupt functions: Data Ready, FIFO flags, pressure thresholds
• Temperature, humidity : HS3003 (datasheet)
• 0-100% relative humidity range
• Humidity accuracy: ±1.5%RH, typical (HS3001, 10 to 90%RH,25°C)
• Temperature sensor accuracy: ±0.1°C, typical
• Up to 14-bit humidity and temperature output data
• DC-DC : P2322
• Regulates input voltage from up to 21V with a minimum of 65% efficiency @minimum load
• More than 85% efficiency @12V

Isi dalam paket :

• 1x Arduino Nano 33 BLE Sense
• 1x OV7675 Camera
• 1x Arduino Tiny Machine Learning Shield
• 1x USB A to Micro USB Cable

Tutorial :
Tiny ML in interactive spaces Arduino X K-WAY Challenge Project
These projects from CMU incorporate the Arduino Nano 33 BLE Sense in clever ways