Overview

IotDevicesPy is a Python lib for controlling IOT devices that contains following packages.

• IotLib - base classes for general IOT components and devices
• LegoLib - classes wrapping Lego MoveHub and components
• CameraLib - encapsulates cameras (Pi and OpenCV) used for video streaming and face tracking

Getting Started

1. download/clone the respository. These steps assume that the code is under /home/pi/IotDevicesPy.
2. Install dependencies.
   • without video streaming
     • LegoLib: pylgbst (see https://github.com/undera/pylgbst for more dependencies)
   • with video streaming and face tracking on Raspberry Pi
     • CameraLib: python3-opencv, dlib, picamera (opencv depends on numpy, libatlas-base-dev)
     • examples: flask
   • with video streaming and face tracking on Windows
     • python-opencv, dlib (see Notes below)
     • pip install Flask
3. Install the IotDevices packages.

• Raspberry Pi

        cd IotDevicesPy
        sudo pip3 install .

• Windows

        cd IotDevicesPy
        pip install .

4. Run samples under examples folder. See Examples section for details. Note that 'sudo' is needed to run in Raspberry Pi.

Details

IotLib

Python base classes to support IOT devices.

• IotNode - the base class for all IOT nodes
• IotMotor - the base class for motors
• IotEncodedMotor - base class for encoded motors
• IotDistanceSensor - base class for sensor measuring distance
• IotMobileBot - base class that implements main functions for a mobile bot

CameraLib

Classes to support camera and face tracking.

LegoLib

Classes to control Boost componnts. LegoLib extends the classes defined in IotLib.

• LegoMoveHub - connect to Move Hub and sending commands
• LegoMotor - implements controls to lego motor
• BoostBot - implements controls to a Boost bot (like Vernie)
• BoostCommandBot - extends BoostBot to support using text command

Examples

Sample codes are provided under the examples folder. Also see https://github.com/rphuang/LegoBoostCmdPy for more usages. Note that 'sudo' is needed to run in Raspberry Pi.

• legoBoostSample-auto.py - simple code to run Boost in auto modes (follow, wander)
• legoBoostSample-bot.py - simple code to run Boost using BoostBot class
• legoBoostSample-cmd.py - simple code to run Boost using BoostCommandBot class
• legoBoostSample-stream.py - simple code to run Boost along with video streaming from a Pi Zero W as in this picture. Note that the resolution is 320x240 due to Pi Zero limitation.
• streamingService.py - sample video streaming service using flask
• streamSample-pi.py - simple code to stream on RasPi
• streamSample-win.py - simple code to stream on Windows

Notes, Issues

• MoveHub automatically switch off a lot immediately after connected on Raspberry Pi Zero (for both pygatt and bluepy). However, that never happened for Windows and Raspberry Pi 4.
• Blue tooth package used for testing
  • Windows: bleak
  • Raspberry Pi: pygatt
• install opencv on Windows with Visual Studio Community 2022 and Python 3.10
  1. install "Desktop development with C++" in Visual Studio
  2. install cmake: pip install CMake
  3. download dlib from https://github.com/davisking/dlib and unzip
  4. edit numpy.h (see davisking/dlib#2463)
  5. build & install dlib from the unzipped folder: py setup.py install
  6. install numpy & opencv: pip install numpy opencv-python
• Maybe it's better to replace Vernie's head with Raspberry Pi 4 for better stability and performance.

ToDos

• RasPi components and devides
• Kasa smart plug
• Tello drone
• Leverage Android phone's sensors