• RoboCar project documentation
  • 1 Project RoboCar
    • 1.1 How to join us
    • 1.2 FAQ
  • 2 Building and operate a donkeycar
    • 2.1 Reason for changing the plattform
      • 2.1.1 Selecting a RC car
    • 2.2 What to do
    • 2.3 Rewiring the car
      • 2.3.1 Parts list
      • 2.3.2 Why not using PCA9685
    • 2.4 Folders created during installation of host PC
    • 2.5 Calibrate the car
      • 2.5.1 Setting WIFI up for the raspberry
      • 2.5.2 Steering Calibration
      • 2.5.3 Throttle Calibration
      • 2.5.4 Config.py after calibration
    • 2.6 Drive your car with web interface
      • 2.6.1 Preparation on Mac
      • 2.6.2 Preparation on Raspberry
      • 2.6.3 Open control in web browser on Mac
    • 2.7 Connecting bluetooth controller
      • 2.7.1 Connect your bluetooth controller to the raspberry pi.
    • 2.8 Start driving or training
      • 2.8.1 Create movie from tub data
      • 2.8.2 Options of manage.py
    • 2.9 Tips on how to drive during training
      • 2.9.1 How to edit recorded data
    • 2.10 Mange recorded data
      • 2.10.1 Archiving data
      • 2.10.2 Delete data
      • 2.10.3 Copy files from raspi to mac
      • 2.10.4 How big is the tub folder
    • 2.11 Train neural network
      • 2.11.1 Add augmentation
      • 2.11.2 Transfer learning
    • 2.12 Run car on neural net
      • 2.12.1 change model
    • 2.13 See Rainer’s description
    • 2.14 Build pre processing pipe line
      • 2.14.1 Install openCV on the raspi
    • 2.15 bluetooth key mapping
    • 2.16 Load h5 model and inspect
  • 3 Wireless connections Raspi, Controller, Mac
    • 3.1 Purpose of connections
    • 3.2 Setting up of the wireless connections
      • 3.2.1 Bluetooth connection
      • 3.2.2 Wifi connection
  • 4 Donkey Command-line Utilities
    • 4.1 Create Car
    • 4.2 Update Car
    • 4.3 Find Car
    • 4.4 Calibrate Car
    • 4.5 Clean data in Tub
    • 4.6 Make Movie from Tub
    • 4.7 Check Tub
    • 4.8 Histogram
    • 4.9 Plot Predictions
    • 4.10 Simulation Server
    • 4.11 Colab training
      • 4.11.1 Blog: Train donkeycar with GPU support in Colab
  • 5 Donkeycar software analysis
    • 5.1 How to add parts to a vehicle
      • 5.1.1 Function V.start()
      • 5.1.2 Function V.update_parts()
      • 5.1.3 Function V.add
      • 5.1.4 Concept of threads vs non-threaded parts
    • 5.2 Structure of manage.py
      • 5.2.1 Creating manage.py
      • 5.2.2 Starting manage.py
      • 5.2.3 Parsing the command line parameters
      • 5.2.4 Function drive
    • 5.3 Locate modules on raspi which are being imported by manage.py
    • 5.4 Assignement of Logitech 710 controler
    • 5.5 Assignment of WiiU controller sticks to action
    • 5.6 Limit max velocity
      • 5.6.1 Limit max velocity for joystick input
      • 5.6.2 Limit max velocity for ai pilot
    • 5.7 Add LED to indicate recording mode
      • 5.7.1 Wiring of LED to PWM/Servo HAT
    • 5.8 Change default recording status to false
  • 6 Installation of Tawn Kramers donkeycar version (By Rainer Bareiß)
    • 6.1 Environment
    • 6.2 sudo raspi-config
    • 6.3 etc/wpa_supllicant/wpa_supplicant.conf
    • 6.4 installs
    • 6.5 install donkey on pi
    • 6.6 5. config.piy
    • 6.7 bluetooth
    • 6.8 8. wii u controller
    • 6.9 autostart donkeycar
    • 6.10 Steering Calibration
    • 6.11 drive
  • 7 Driving & Training following Tawn Kramer (By Rainer Bareiß)
    • 7.1 Environment
    • 7.2 Read data from pi
    • 7.3 Check data raw & prepared
    • 7.4 3. Clean data
    • 7.5 Make movie
    • 7.6 tubhist
    • 7.7 Train Data
    • 7.8 Plot data against model
    • 7.9 Run in simulator
    • 7.10 Copy model to pi
    • 7.11 Run model on pi
    • 7.12 11. Watch pi camera
    • 7.13 Reinfocement Learning
  • 8 RoboCar Logbook
    • 8.1 15-02-2018
      • 8.1.1 donkey findcar
    • 8.2 Build sunfounder car 17-02-2018
      • 8.2.1 install software 18-02-2018
    • 8.3 Viewing a list of available Python versions
    • 8.4 Run calibration
    • 8.5 MJPG-streamer Installation
      • 8.5.1 Testing
  • 9 Get on the Road!
    • 9.1 install carnd-term1
    • 9.2 Install Xbox 360 controller
    • 9.3 Implement Xbox controller as input in client_App.py
      • 9.3.1 Make steering proportional to remote control lever position
    • 9.4 Get IP adress of raspi in shack
    • 9.5 Store snapshots of video stream on local computer
      • 9.5.1 Jie Hou’s alternative
  • 10 Resources
    • 10.1 Meetups
      • 10.1.1 Esslinger Makerspace Projekt: Autonomen RoboCar bauen
      • 10.1.2 Autonomous Mobility Berlin
    • 10.2 RoboCar projects
      • 10.2.1 DIY RoboCars
      • 10.2.2 Donkey car
      • 10.2.3 Sunfounder Smart Video Car Kit for Raspberry Pi with Android App
      • 10.2.4 Kuman Professional WIFI Smart Robot Model Car Kit Videokamera for Raspberry Pi 3
    • 10.3 Hardware
      • 10.3.1 KOMPONENTENLISTE FÜR EIN FERNGESTEUERTES ROBOTER AUTO
      • 10.3.2 Motor control
    • 10.4 Simulators
      • 10.4.1 Donkey Simulator
    • 10.5 Software
      • 10.5.1 Udacitiy open source SDC
      • 10.5.2 connect raspberry with xbox controller
      • 10.5.3 RASPBERRY PI RC CONTROL
      • 10.5.4 Python class to support xbox 360 controller under Linux on RaspberryPi
      • 10.5.5 xbox driver for mac
      • 10.5.6 Konzept für softwareschichten
  • 11 Experiences gathered
    • 11.1 Install donkey car on Mac
    • 11.2 Donkey car simulator
  • 12 Theory
    • 12.1 Wiring of PCA9685 with L298N H-bridge
    • 12.2 sensor fusion of LIDAR and camera data
  • 13 Useful links
    • 13.1 Donkey car
      • 13.1.1 Videos for donkey car
    • 13.2 Parts to build a RoboCar
    • 13.3 Little setup helpers
    • 13.4 Meetups in Stuttgart area
    • 13.5 RoboCar projects
    • 13.6 Write ups on Donkeycar projects
    • 13.7 Other stuff
  • Related files on Github

RoboCar project documentation

Chapter 13 Useful links

Links I found useful at some point

13.1 Donkey car

Here are links to the donkey car project

  • A great buying guide for RC-cars
  • Calibrate your car
  • Drive your car
  • Train an autopilot with Keras
  • http://docs.donkeycar.com/guide/simulator/
  • Controller Parts
  • donkey Car FAQ
  • Donkey car CAD F360 file

13.1.1 Videos for donkey car

Tawn Kramer on how to add parts (like a camera) to the vehicle. Great step by step intro to the Vehicle class of the framework.
- Donkeycar Parts Overview Pt 1/2
- Donkeycar Parts Overview Pt 2/2

Another video by Tawn Kramer gives tips on how to drive during training

Driving Tips to Train your Autonomous End-to-End NN Driver

This is the official Donkey Car Assembly video. It includes using options like the Sombrero board and The NVIDIA Jetson Nano

  • Donkey Car Assembly Video

13.2 Parts to build a RoboCar

Links to hardware

Reely Dart 2.0 Brushed

PWM SERVO HAT

Adafruit 16-Channel PWM/Servo HAT & Bonnet for Raspberry Pi16 channels of servo-bust

Adafruit 16 Channel Servo Driver with Raspberry Pi

Raspberry Pi

Raspbery Pi camera

DC/DC converter for raspi

WiiU controller

Tamiya Stecker Power wires

Template for 3D printing a plate for raspi and DC/DC converter

13.3 Little setup helpers

Here are links to instructions on how to get things done

SETTING WIFI UP VIA THE COMMAND LINE

Install XBox driver on mac

Python Xbox Controller Mac

find IP adress in shack

Self-driving car in a simulator with a tiny neural network by xslittlegrass

Github of Detroit-Autonomous-Vehicle-Group

HOW TO TETHER RASPBERRY PI TO ANDROID DEVICE HOTSPOT

File Sharing From a Mac to Your Raspberry Pi

SFTP on Raspbery

13.4 Meetups in Stuttgart area

  • Esslinger Makerspace Projekt: Autonomen RoboCar bauen In der Konzeption für die neue/umgebaute Stadtbücherei ist ein Makerspace vorgesehen. Diese tolle Idee sollten wir unterstützen! Und als erstes Projekt entwickeln wir RoboCars die autonom einen Parkour bewältigen können. Ob ihr euer eigenes RoboCar entwickeln wollt, oder lieber ein kleines Team bilden wollt, hier seit ihr richtig. Mal schauen welche Hard- bzw. Softwarekonzepte euch einfallen. Zur Eröffnung des Makerspace Esslingen können wir dann zeigen was wir erreicht haben.

  • Connected Autonomous Driving This Meetup is for autonomous driving enthusiasts, self-driving car engineers and those who want to create a seamless connected user experience of mobility, information and services in cities which fosters smart cities & measurable/scalable solutions saving our climate. We are a Homebrew computer club for autonomous mobility* and connect schools, creatives, engineers, startups, companies, politics for exchange & projects to create a seamless connected user experience of mobility, information and services in cities.

13.5 RoboCar projects

  • DonkeyCar

  • DIY RoboCars website

  • Sunfounder RPI car

  • AWS DeepRacer

13.6 Write ups on Donkeycar projects

Compact Donkey Car by Oliver Wannenwetsch Learning to Drive Smoothly in Minutes Reinforcement Learning on a Small Racing Car

HOW TO SET UP A ROBOCAR PLATFORM WITH A REMOTE CONTROL UNIT by itemis

13.7 Other stuff

ESC explained

LED on The Adafruit 16-Channel 12-bit PWM/Servo HAT

Mount a Raspberry Pi SD card on a Mac (read-only) with osxfuse and ext4fuse

Dimming LEDs with PWM/Servo HAT without the need of an restistor in chapter “Python Wiring”

cross reference headings in bookdown {#AssignStickToAction} 5.5