Features: Main module supporting AI acceleration calculation, multimedia and various IoT sensors are integrated into the base board. Min. 5-inch TFT LCD with min. 800x480 resolution and 8M pixel high resolution camera. Digital microphones and speakers support cloud-based speech recognition and audio playback. Min. 4 dedicated expansion interfaces support various IoT sensor modules Soda OS, the exclusive AIoT operating system, and Pop library. A dedicated web browser-based learning environment for training Python 3 and C/C++ Educational contents for IoT sensor control, multimedia and AI.
Training Contents:
Introduction to AIoT Home:
1. Configuration and Practice Environment of AIoT Home Python and Linux 101.
2. IoT Application Technology Sensor Control: File and DB-Based Data Persistence Audio Recording and Playback Google Text-to-Speech Converter.
3. Google Assistant and User Device Actions Camera and Sensor Applications
AI Technology:
1. Numpy for Fast Multidimensional Matrix Operations Pandas for Time Series and Tabular Data Analysis Matplotlib for Data Visualization
2. Supervised and Unsupervised Learning.
3. Theory & Practice for Pop.AI-based Linear and Logistic Regression Algorithm, Perceptron, ANN, DNN, and CNN & OpenAI DQN-based Reinforcement Learning Understanding TensorFlow.
Software Specifications:
Soda OS:
Linux Kernel: 4.19
Desktop: X-Server, Openbox, LightDM, Tint2, blueman, network-manager, conky
CLI: Zsh, Tmux, Peco, powerlevel9k thema, Powerline fonts IDE: Visual Studio Code, NeoVim, Geany
Data Science & AI: Python3, Numpy, Matplotlib, sympy, Pandas, Seaborn, Scipy, Gym Scikit-learn, Tensorflow, Kerast
Pop Library:
Output Object (C/C++, Python3): Led, Laser, Buzzer, Relay, RGBLed, DCMotor, StepMotor, OLed, PiezoBuzzer, PixelDisplay, TextLCD, FND, Led Bar
Input Object (C/C++, Python3): Switch, Touch, Reed, LimitSwitch, Mercury, Knock, Tilt, Opto, Pir, Flame LineTrace, TempHumi, UltraSonic, Shock, Sound, Potentiometer, CdS, SoilMoisture, Thermistor, Temperature, Gas, Dust, Psd. Gesture
AI (Python3): Linear Regression, Logistic Regression, Perceptron, ANN, DNN, CNN, DQN
Hardware Specifications:
Main Module:
CPU: 6-core NVIDIA Carmel ARM v8.2 64-bit; 6MB L2 + 4MB L3; Max Freq: 2- core@1900MHz, 4/6-core@1400Mhz
GPU: 384-core NVIDIA VoltaTM GPU with 48 Tensor Cores; Max Freq: 1100MHz Memory: 8GB 128-bit LPDDR4x@ 1600MHz
Video Encoder: 2x464MP/sec(HEVC); 2x4k@30(HEVC); Video Decoder: 2x690MP/sec(HEVC), 2x4k@ 60(HEVC) CSI Camera: Up to 6 cameras (36 via virtual channels);
Connectivity: Dual Band Wireless Wi-Fi 2GHz/5GHz Band, 867Mbps, 802.11ac; Bluetooth 4.2; 10/100/1000 Base-T Ethernet
Display: 2 multi-mode DP 1.4/eDP 1.4/HDMI 2.0 USB: 4x USB 3.0, USB 2.0 Micro-B
Base Board:
Camera:
Image Sensor:
Video: 1080p30, 720p60 and 640x480p90 Linux Integration: V4L2 Driver Available Sound:
Sound IC: WM8960
Channel: Input 2ch, Output 2ch 2CH Microphone
Stereo Speaker Servo Motor:
Dead Zone Width: 5?sec Neutral Location: 1500us Interface: I2C
TEXT LCD:
Interface: GPIO LED:
(1x3) Group x 3EA: Light Display Stand Light Display 1EA Interface: GPIO
RGB LED:
Size: 5pi
Wavelength: RED (630 nm), GREEN (525 nm), BLUE (430 nm) Interface: I2C
PIEZO:
Rated Current: Max 30mA Interface: GPIO
FAN:
Size: Min. 30 X 30mm Speed: Min. 900RPM Supply Voltage: 5V LED BAR x 2EA:
Size: 20 X 10mm
Humidity & Temperature Sensor:
Humidity Resolution: 12bit (0.04%RH), 8bit (0.7%RH) Humidity Accuracy: +-3%RH
Temperature Resolution: 14bit (0.01C), 12bit (0.04C) Temperature Accuracy: +-4°C
Interface: I2C
Supply Voltage: 3.3V PIR Sensor:
Transmittance: ?75%
Signal Output [Vp-p]: ?3500mV Sensitivity: ?3300V/W
Detect: ?1.4X108cm Detecting Distance: 10~80cm Interface: GPIO
Supply Voltage: 3.3V DUST Sensor:
Based on Laser Scattering Technology Interface: HDMI
Illuminance Sensor:
Sensor: CdS Interface: ADC GAS Sensor:
Sensor Block2: +5V, +3.3V, GND, I2C, ADC, GPIO Sensor Block3: +5V, +3.3V, GND, SPI, GPIO Sensor Block4: +5V, +3.3V, GND, ADC, GPIO
Sensor Pack:
Flame Module:
Sensing Range: 60 Degree
Eco Sensor Module:
Light Sensor: Illuminance to digital converter.
