Skip to content

Robot and controller code for a replica robot

License

Notifications You must be signed in to change notification settings

HAstral/walle-replica

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

52 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

GPLv3 License Issues GitHub last commit

Wall-E Robot Replica

Robot and controller code for a Wall-E replica robot. For more information about the robot, visit https://wired.chillibasket.com/3d-printed-wall-e/



1. Arduino Code (wall-e)

Main program to control the motors and servos of the robot. Features include:

  1. An animation queue, keeping track of the next servo movements the robot needs to perform.
  2. A random movement generator, allowing the robot to autonomously move and appear animated.
  3. Velocity control of all servo motors, facilitating smooth accelerations and decelerations.
  4. Non-blocking serial parsing, allowing the movements of the robot to be remote controlled.
  5. Battery level monitoring using a potential divider circuit.

2. Raspberry Pi Web Server (web_interface)

The web interface is programmed in Python and uses Flask to generate a server. The Raspberry Pi is connected via USB to the Arduino micro-controller. The main features are:

  1. A JavaScript joystick, with which the movement of the robot can easily be controller.
  2. Manual control of all the servo motors.
  3. A list of movement animations which can be performed by the robot.
  4. A list of sounds which can be played.
  5. Settings page, where motor parameters, sound volume, and video options can be modified.
  6. Gamepad support; on any modern browsers, you can use a connected Xbox of PlayStation controller to control the robot.
  7. A simple login page to prevent everyone from having access to the controls (note: this is not a full access control system, please don't use this web interface on untrusted/public networks)

Image of the web interface and robot



Setup Instructions

1. Arduino

[a] Basic Installation

  1. Ensure that the wiring of the electronics matches the diagram shown below.
  2. Download/clone the folder "wall-e" from the GitHub repository.
  3. Open wall-e.ino in the Arduino IDE; the files animations.ino, MotorController.hpp and Queue.hpp should automatically open on separate tabs of the IDE as well.
  4. Install the Adafruit_PWMServoDriver.h library
    1. Go to Sketch -> Include Library -> Manage Libraries...
    2. Search for Adafruit Servo.
    3. Install latest version of the library.
  5. Connect to the computer to the micro-controller with a USB cable. Ensure that the correct Board and Port are selected in the Tools menu.
  6. Upload the sketch to the micro-controller.

Diagram showing the wiring of the robot's electronic components


[b] Testing the Main Program

  1. Once the sketch has been uploaded to the Arduino, power on the 12V battery while the micro-controller is still connected to the computer.
  2. Open the Serial Monitor (button in top-right of Arduino IDE). Set the baud rate to 115200.
  3. To control the movement of the robot, send the characters 'w', 'a', 's' or 'd' to move forward, left, back or right respectively. Send 'q' to stop all movement.
  4. To move the head, send the characters 'j', 'l', 'i' or 'k' to tilt the head left or right and the eyes upwards or downwards. At this stage, the servos may try to move further than they should and may look uncoordinated. This will be solved by performing the servo motor calibration steps below.

[c] Servo Motor Calibration

  1. Download/clone the folder "wall-e_calibration" from the GitHub repository
  2. Open wall-e_calibration.ino in the Arduino IDE.
  3. Upload the sketch to the micro-controller, and open the serial monitor and set the baud rate to 115200.
  4. The sketch is used to calibrate the maximum and minimum PWM pulse lengths required to move each servo motor across its desired range of motion. The standard LOW and HIGH positions of each of the servos can be seen on diagrams on my website.
  5. When starting the sketch and opening the serial monitor, the a message should appear after 2-3 seconds, saying that it is ready to calibrate the LOW position of the first servo motor (the head rotation).
  6. Send the character 'a' and 'd' to move the motor backwards and forwards by -10 and +10. For finer control, use the characters 'z' and 'c' to move the motor by -1 and +1.
  7. Once the motor is position in the correct position, send the character 'n' to proceed to the calibration step. It will move on to the HIGH position of the same servo, after which the process will repeat for each of the 7 servos in the robot.
  8. When all joints are calibrated, the sketch will output an array containing the calibration values to the serial monitor.
  9. Copy the array, and paste it into lines 144 to 150 of the program wall-e.ino. The array should look similar to this:
    int preset[][2] =  {{410,120},  // head rotation
                        {532,178},  // neck top
                        {120,310},  // neck bottom
                        {465,271},  // eye right
                        {278,479},  // eye left
                        {340,135},  // arm left
                        {150,360}}; // arm right

[d] Battery Level Detection (Optional)

When using batteries to power the robot, it is important to keep track of how much power is left. Some batteries may break if they are over-discharged, and the SD card of the Raspberry Pi may become corrupted if not enough power is delivered.

  1. To use the battery level detection feature on the Arduino, connect the following resistors and wiring as shown in the image below. The resistors (potential divider) reduce the 12V voltage down to a value below 5V, which is safe for the Arduino to measure using its analogue pins. The recommended resistor values are R1 = 100kΩ and R2 = 47kΩ.
  2. Uncomment line 54 #define BAT_L in the main Arduino sketch wall-e.ino.
  3. If you are using different resistor values, change the value of the potential divider gain factor on line 54 of the sketch, according to the formula: POT_DIV = R2 / (R1 + R2).
  4. The program should now automatically check the battery level every 10 seconds, and this level will be shown on the Raspberry Pi web-interface in the "Status" section.


Diagram showing the wiring of the battery level detection circuit


[e] oLed Display (Optional) (Contributed by: hpkevertje)

It is possible to integrate a small oLED display which will show the battery level of the robot on the front battery indicator panel. This feature requires the battery level detection circuit in the previous section to be enabled, and the screen will update every time the battery level is calculated. This function uses the u8g2 display library in page mode; on the Arduino UNO you may get a warning that the memory usage is high, but this warning can be ignored.

  1. To use the oLed display feature on the Arduino, connect an i2c oLed display on the i2c bus on the servo motor module (see diagram).
  2. Install the U8g2 library in the Arduino library manager:
    1. Go to Sketch -> Include Library -> Manage Libraries...
    2. Search for U8gt. The library publisher is "oliver".
    3. Install latest version of the library.
  3. Uncomment line 74 #define OLED in the main Arduino sketch wall-e.ino.
  4. If you are using a different display that is supported by the library, you can change the constructor on line 78 as documented on the library reference page. The default is for an SH1106_128X64_NONAME display.


Diagram showing the wiring of the oLed display


[f] Adding your own Servo Animations (Optional)

My code comes with two animations which replicate scenes from the movie; the eye movement Wall-E does when booting-up, and a sequence of motions as Wall-E inquisitively looks around. From version 2.7 of the code, I've now made it easier to add your own servo motor animations so that you can make your Wall-E do other movements...

  1. Open up the file animations.ino, which is located in the same folder as the main Arduino sketch.
  2. Each animation command consists of the positions you want each of the servo motors to move to, and the amount of time the animation should wait until moving on to the next instruction.
  3. You can add a new animation by inserting an extra case section into the switch statement. You should slot your extra code into the space above the default section. For example:
    case 3:
            // --- Title of your new motion sequence ---
            //          time,head,necT,necB,eyeR,eyeL,armL,armR
            queue.push({  12,  48,  40,   0,  35,  45,  60,  59});
            queue.push({1500,  48,  40,  20, 100,   0,  80,  80});
            // Add as many additional movements here as you need to complete the animation
            // queue.push({time, head rotation, neck top, neck bottom, eye right, eye left, arm left, arm right})
            break;
  4. The time needs to be a number in milliseconds (for example, 3.5 seconds = 3500)
  5. The servo motor position commands need to be an integer number between 0 to 100, where 0 = LOW and 100 = HIGH servo position as calibrated in the wall-e_calibration.ino sketch.
  6. If you want to disable a motor for a specific move, you can use -1.


2. Raspberry Pi Web Server

[a] Basic Installation

  1. Setup the Raspberry Pi to run the latest version of Raspberry Pi OS (Raspbian) - Full. The setup instructions can be found on the Raspberry Pi website.
  2. Open the command line terminal on the Raspberry Pi.
  3. Ensure that the package list has been updated (this may take some time): sudo apt update
  4. Install Flask - this is a Python framework used to create web servers:
    1. Ensure that pip is installed: sudo apt install python3-pip
    2. Install Flask and its dependencies: sudo pip3 install flask
  5. (Optional) The Full version of Raspbian includes these packages by default, but if you are using a different OS (for example the Lite version), you will need to run these commands:
    sudo apt install git libsdl1.2 libsdl-mixer1.2
    sudo pip3 install pygame pyserial
  6. Clone repository into the home directory of the Raspberry Pi:
    cd ~
    git clone https://github.com/chillibasket/walle-replica.git
  7. Set the web server password:
    1. Open app.py: nano ~/walle-replica/web_interface/app.py
    2. On line 26 of app.py where is says put_password_here, insert the password you want to use for the web interface.
  8. (Optional) Change the default audio directory and location of the script used to start/stop the video stream.
    1. If you followed the steps above exactly, there is no need to do this. However, if you want to move the web-interface files to a different directory on the Raspberry Pi, you will need to change the location where the program will look for the audio files.
    2. On line 29 of app.py, type the directory where the audio files are located. Ensure that the directory location ends with a forward slash: /.
    3. On line 28 of app.py, the location of the script used to start and stop the video camera stream can be modified.
  9. Connect to the Arduino/micro-controller:
    1. Plug the Arduino/micro-controller into the USB port of the Raspberry Pi.
    2. If you would like the serial port used by the Arduino to be selected by default in the web-interface, you can set a preferred serial port device in the code. Go to line 27 of app.py and replace the text "ARDUINO" with the name of your device. The name must match the one which appears in the drop-down menu in the "Settings" tab of the web-interface.
    3. To make the interface automatically connect to the Arduino when it starts up, you can change line 31 to autoStartArduino = True
    4. Press CTRL + O to save and CTRL + X to exit the nano editor.

[b] Using the Web Server

  1. To determine the current IP address of the Raspberry Pi on your network, type the command: hostname -I
  2. To start the server: python3 ~/walle-replica/web_interface/app.py
  3. To access the web interface, open a browser on any computer/device on the same network and type in the IP address of the Raspberry Pi, follow by :5000. For example 192.168.1.10:5000
  4. To stop the server press: CTRL + C
  5. To start controlling the robot, you first need to start serial communication with the Arduino. To do this, go to the Settings tab of the web-interface, select the correct serial port from the drop-down list and press on the Reconnect button.

[c] Adding a Camera Stream (Optional)

  1. If you are using the Official Raspberry Pi camera, you will first need to enable the camera in sudo raspi-config. In the config screen which appears, navigate to “Interface Options” > “Camera” > “Enable”.
  2. Install mjpg-streamer - this is used to stream the video to the webserver. The steps below are based on the instructions described here for the CNC JS project. To install the required libraries, run all of the following commands:
    # Update & Install Tools
    sudo apt-get update -y
    sudo apt-get upgrade -y
    sudo apt-get install build-essential git imagemagick libv4l-dev libjpeg-dev cmake -y
    
    # Clone Repo in /tmp
    cd /tmp
    git clone https://github.com/jacksonliam/mjpg-streamer.git
    cd mjpg-streamer/mjpg-streamer-experimental
    
    # Make
    make
    sudo make install
    
  3. Create a new script which will be used to start and stop the camera stream: nano ~/mjpg-streamer.sh
  4. Paste the following code into the script file; you can modify the settings frame rate, quality and resolution to suit the camera you are using:
    #!/bin/bash
    # chmod +x mjpg-streamer.sh
    # Crontab: @reboot /home/pi/mjpg-streamer/mjpg-streamer.sh start
    # Crontab: @reboot /home/pi/mjpg-streamer/mjpg-streamer-experimental/mjpg-streamer.sh start
    
    MJPG_STREAMER_BIN="/usr/local/bin/mjpg_streamer"  # "$(dirname $0)/mjpg_streamer"
    MJPG_STREAMER_WWW="/usr/local/share/mjpg-streamer/www"
    MJPG_STREAMER_LOG_FILE="${0%.*}.log"  # "$(dirname $0)/mjpg-streamer.log"
    RUNNING_CHECK_INTERVAL="2" # how often to check to make sure the server is running (in seconds)
    HANGING_CHECK_INTERVAL="3" # how often to check to make sure the server is not hanging (in seconds)
    
    VIDEO_DEV="/dev/video0"
    FRAME_RATE="5"
    QUALITY="80"
    RESOLUTION="1280x720"  # 160x120 176x144 320x240 352x288 424x240 432x240 640x360 640x480 800x448 800x600 960x544 1280x720 1920x1080 (QVGA, VGA, SVGA, WXGA)   #  lsusb -s 001:006 -v | egrep "Width|Height" # https://www.textfixer.com/tools/alphabetical-order.php  # v4l2-ctl --list-formats-ext  # Show Supported Video Formates
    PORT="8080"
    YUV="true"
    
    ################INPUT_OPTIONS="-r ${RESOLUTION} -d ${VIDEO_DEV} -f ${FRAME_RATE} -q ${QUALITY} -pl 60hz"
    INPUT_OPTIONS="-r ${RESOLUTION} -d ${VIDEO_DEV} -q ${QUALITY} -pl 60hz --every_frame 2"  # Limit Framerate with  "--every_frame ", ( mjpg_streamer --input "input_uvc.so --help" )
    
    
    if [ "${YUV}" == "true" ]; then
        INPUT_OPTIONS+=" -y"
    fi
    
    OUTPUT_OPTIONS="-p ${PORT} -w ${MJPG_STREAMER_WWW}"
    
    # ==========================================================
    function running() {
        if ps aux | grep ${MJPG_STREAMER_BIN} | grep ${VIDEO_DEV} >/dev/null 2>&1; then
            return 0
    
        else
            return 1
    
        fi
    }
    
    function start() {
        if running; then
            echo "[$VIDEO_DEV] already started"
            return 1
        fi
    
        export LD_LIBRARY_PATH="$(dirname $MJPG_STREAMER_BIN):."
    
        echo "Starting: [$VIDEO_DEV] ${MJPG_STREAMER_BIN} -i \"input_uvc.so ${INPUT_OPTIONS}\" -o \"output_http.so ${OUTPUT_OPTIONS}\""
        ${MJPG_STREAMER_BIN} -i "input_uvc.so ${INPUT_OPTIONS}" -o "output_http.so ${OUTPUT_OPTIONS}" >> ${MJPG_STREAMER_LOG_FILE} 2>&1 &
    
        sleep 1
    
        if running; then
            if [ "$1" != "nocheck" ]; then
                check_running & > /dev/null 2>&1 # start the running checking task
                check_hanging & > /dev/null 2>&1 # start the hanging checking task
            fi
    
            echo "[$VIDEO_DEV] started"
            return 0
    
        else
            echo "[$VIDEO_DEV] failed to start"
            return 1
    
        fi
    }
    
    function stop() {
        if ! running; then
            echo "[$VIDEO_DEV] not running"
            return 1
        fi
    
        own_pid=$$
    
        if [ "$1" != "nocheck" ]; then
            # stop the script running check task
            ps aux | grep $0 | grep start | tr -s ' ' | cut -d ' ' -f 2 | grep -v ${own_pid} | xargs -r kill
            sleep 0.5
        fi
    
        # stop the server
        ps aux | grep ${MJPG_STREAMER_BIN} | grep ${VIDEO_DEV} | tr -s ' ' | cut -d ' ' -f 2 | grep -v ${own_pid} | xargs -r kill
    
        echo "[$VIDEO_DEV] stopped"
        return 0
    }
    
    function check_running() {
        echo "[$VIDEO_DEV] starting running check task" >> ${MJPG_STREAMER_LOG_FILE}
    
        while true; do
            sleep ${RUNNING_CHECK_INTERVAL}
    
            if ! running; then
                echo "[$VIDEO_DEV] server stopped, starting" >> ${MJPG_STREAMER_LOG_FILE}
                start nocheck
            fi
        done
    }
    
    function check_hanging() {
        echo "[$VIDEO_DEV] starting hanging check task" >> ${MJPG_STREAMER_LOG_FILE}
    
        while true; do
            sleep ${HANGING_CHECK_INTERVAL}
    
            # treat the "error grabbing frames" case
            if tail -n2 ${MJPG_STREAMER_LOG_FILE} | grep -i "error grabbing frames" > /dev/null; then
                echo "[$VIDEO_DEV] server is hanging, killing" >> ${MJPG_STREAMER_LOG_FILE}
                stop nocheck
            fi
        done
    }
    
    function help() {
        echo "Usage: $0 [start|stop|restart|status]"
        return 0
    }
    
    if [ "$1" == "start" ]; then
        start && exit 0 || exit -1
    
    elif [ "$1" == "stop" ]; then
        stop && exit 0 || exit -1
    
    elif [ "$1" == "restart" ]; then
        stop && sleep 1
        start && exit 0 || exit -1
    
    elif [ "$1" == "status" ]; then
        if running; then
            echo "[$VIDEO_DEV] running"
            exit 0
    
        else
            echo "[$VIDEO_DEV] stopped"
            exit 1
    
        fi
    
    else
        help
    
    fi
    
  5. Press CTRL + O to save and CTRL + X to exit the nano editor.
  6. Make sure that the manager script you created has the correct name and is in the correct directory: /home/pi/mjpg-streamer.sh. If you want the save the script in a different location, you need to update line 22 of app.py.
  7. To make the script executable by the web-server, run this command in the terminal: chmod +x /home/pi/mjpg-streamer.sh
  8. If you want the camera to automatically startup when you open the web-interface you can change line 32 of app.py to autoStartCamera = True

[d] Automatically start Server on Boot (Optional, but recommended)

  1. Create a .service file which is used to start the web interface: nano ~/walle.service
  2. Paste the following text into the file:
    [Unit]
    Description=Start Wall-E Web Interface
    After=network.target
    
    [Service]
    WorkingDirectory=/home/pi/walle-replica/web_interface
    ExecStart=/usr/bin/python3 app.py
    Restart=always
    StandardOutput=syslog
    StandardError=syslog
    SyslogIdentifier=walle
    User=pi
    
    [Install]
    WantedBy=multi-user.target
    
  3. Press CTRL + O to save and CTRL + X to exit the nano editor.
  4. Copy this file into the startup directory using the command: sudo cp ~/walle.service /etc/systemd/system/walle.service
  5. To enable auto-start, use the following command: sudo systemctl enable walle.service
  6. The web interface should now automatically start when the Raspberry Pi is turned on. You can also manually start and stop the service using the commands: sudo systemctl start walle.service and sudo systemctl stop walle.service

[e] Adding new Sounds (Optional)

  1. By default the Raspberry should automatically select whether to output audio to the HDMI port or the headphone jack. However, you can ensure that it always uses the headphone jack with the following command: amixer cset numid=3 1
  2. Make sure that all the sound files you want to use are of type *.ogg. Most music/sound editors should be able to convert the sound file to this format.
  3. Change the file name so that it has the following format: [group name]_[file name]_[length in milliseconds].ogg. For example: voice_eva_1200.ogg. In the web-interface, the audio files will be grouped using the "group name" and sorted alphabetically.
  4. Upload the sound file to Raspberry Pi in the following folder: ~/walle-replica/web_interface/static/sounds/
  5. All the files should appear in the web interface when you reload the page. If the files do not appear, you may need to change the privileges required to access the folder: sudo chmod -R 755 ~/walle-replica/web_interface/static/sounds

[f] Set up the Raspberry Pi as a WiFi hotspot (Optional)

If you would like to control the robot outdoors or at conventions, there may not be any safe WiFi networks you can connect to. To overcome this issue and eliminate the need for any external networking equipment, the Raspberry Pi can broadcast its own WiFi network. You can then connect the computer/phone/tablet you are using to control the robot directly to this network.

To set up the WiFi hotspot, we will use the RaspAP project which takes care of all the configuration and tools to get the system working. The following instructions are based on their quick installation guide:

  1. Update Raspian, the kernel and firmware (and then reboot):
    sudo apt-get update
    sudo apt-get dist-upgrade
    sudo reboot
    
  2. Ensure that you have set the correct WiFi country in raspi-config’s Localisation Options: sudo raspi-config
  3. Run the quick installer: curl -sL https://install.raspap.com | bash
    1. For the first few yes/no prompts which will appear during the install, type “y” (yes) to accept all of the recommended settings. The final two prompts (Ad Blocking and the next one) are not required so you can type “n” (no) for those.
  4. Reboot the Raspberry Pi again to implement the changes: sudo reboot
  5. Now the Raspberry Pi should be broadcasting a WiFi network with the following details:
    1. SSID (wifi name): raspi-webgui
    2. Password: ChangeMe
  6. After connecting to the WiFi network from a your computer, phone or tablet, the Wall-E web-interface can be opened by typing this address into your browser: http://10.3.141.1:5000
  7. (Recommended) To change the WiFi name and password, go to the WiFi configuration webpage at: http://10.3.141.1. The default username is admin and password is secret.
    1. Click on “Hotspot” in the left sidebar. In the “Basic” tab you can change the WiFi network name, while the WiFi password can be changed in the “Security” tab.
    2. To change the admin password for the interface used to manage the WiFi settings, click on the “Admin” icon in the top-right of the interface.


Changelog

31st October 2021 (Version 2.92)

  1. Added options in app.py to automatically connect to the Arduino and to start the camera stream when the web-interface is opened for the first time.

30th October 2021 (Version 2.91)

  1. Changed camera stream port from 8081 to 8080 inline with changes made to the start/stop script.
  2. Improved the camera setup instructions, listing all steps explicitly rather than linking to an external repository.

29th May 2021 (Version 2.9)

  1. Improved code commenting and variable/parameter naming to make it easier to understand.
  2. Fixed bugs related to motor deadzone and trim parameters.

21st March 2021 (Version 2.8)

  1. Updated instructions of how to use an oLed display.

19th December 2020

  1. Updated instructions of how to set up a WiFi hotspot on the Raspberry Pi.
  2. Added offline Lato font files to prevent errors when running the web-interface while disconnected from the internet.

7th August 2020 (Version 2.7)

  1. Added a soft servo start function to the main Wall-E sketch and the servo calibration sketch. This prevents the servos from jumping as violently on startup.
  2. Changed the data-type of the animation queue to reduce the amount of dynamic memory required. Also updated the Queue class so that the buffer memory can be declared globally; this means the compiler can keep track of how much memory the queue actually uses.
  3. Moved the preset servo animations into a separate file, to make it easier to add your own animations.

20th June 2020

  1. Minor bug fixes of the Queue.hpp and MotoController.hpp classes.
  2. Updated commenting of code to make it more consistent.

16th February 2020

  1. Added gamepad support; any controller can now be used (such as the Xbox or PlayStation controllers) to puppet the robot.
  2. Improved serial port handling; in the settings tab, all of the available serial ports are listed in a drop-down menu.
  3. New status indicators are now shown below the video stream. They visualise whether the Arduino and/or gamepad are connected.
  4. Added support for battery level detection in both the Arduino code and the Raspberry Pi web interface.
  5. Fixed the bug where the virtual joystick no longer worked properly when the window was resized.
  6. Plus many other bug fixes!

25th January 2020

  1. Restructured the web-interface to make it display more nicely on mobile devices; fancy icons are now used!
  2. Added manual servo control - this allows individual servos to be controlled directly from the web-interface.
  3. Improved error handling - error messages are now consistently displayed in a pop-up at the bottom of the page.
  4. Added instructions about how to automatically start server on boot to the Readme.

31st October 2019

  1. Fixed some bugs related to sound playback.
  2. Added 2 sample sound files, which ensures that the sound directory is included in the git files.

30th October 2019

  1. Added wall-e_calibration.ino, with which the maximum and minimum pulse widths of the servo motors can be calibrated.
  2. Updated wall-e.ino to use relative coordinates rather than absolute servo pulse widths for the animation presets. This allows the servo calibration data to be used to ensure the movements are the same on all variants of the robot.

About

Robot and controller code for a replica robot

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 76.5%
  • C++ 11.9%
  • HTML 5.8%
  • Python 4.6%
  • CSS 1.2%