Rover IoT | Intel IoT Road Show 2015

Hi friends,
my name is Douglas Esteves and I'm a enthusiastic with the resources in the Intel Edison .

My friend Gilvan Vieira and I developed a project in Event Intel IoT Road Show 2015 (November 6-7) in São Paulo, Brazil.

The idea the project is to control the robot in a different way, using a Myo armband to control wirelessly the robot movements over the internet.

tips : Step webcam configuration used this link https://github.com/drejkim/edi-cam

Equipaments used:

• Intel Edison
• Rover 5 chassis
• Drive Motor L298H
• Batery Lipo 220mAh
• Jumpers
• Router (internet)
• Myo Armband Webcam

Turn On your Intel Edison for procediments basic :

Conect cable micro USB.

Open Terminal Linux. (gnome-terminal my laptop)

gnome-terminal 

Install screen (Debian $apt-get install screen)

 screen /dev/ttyUSB0 115200 

Press Enter case your screen is black

User : root

Password: "null"

Configure IP.

configure_edison --wifi

select your wifi and enter password.

after write command

 ifconfig

About this link Edi-cam

Use webcam more Intel Edison, Node.Js and WebSockets. (Does not transmit song) listening for the incoming video stream via HTTPm ffmpeg...

Use webcam UVC-Compatible .

Configure repository

vi /etc/opkg/base-feeds.conf 

include

src/gz all  http://repo.opkg.net/edison/repo/all

src/gz edison  http://repo.opkg.net/edison/repo/edison

src/gz core2-32 http://repo.opkg.net/edison/repo/core2-32

before opkg update

opkg install git 

git clone

 git clone  https://github.com/drejkim/edi-cam>

example for link edi-cam Check whether or no the UVC drive is installed.

find /lib/modules/* -name 'uvc'

/lib/modules/3.10.17-poky-edison+/kernel/drivers/media/usb/uvc 

lsmod | grep uvc <br>

uvcvideo 71516 0 

videobuf2_vmalloc 13003 1 

uvcvideo 
videobuf2_core 37707 1uvcvideo 

Install ffmeg

cd /edi-cam/bin

./install_ffmpeg.sh 

Install Nose.js

cd ../web/server

npm install 

modify file.

vi web/client/index.html 

var wsUrl = 'ws://ip:8084'

I used this link (Tank's Esther Jun Kim) : https://github.com/drejkim/edi-cam

Chassi Rover 5

Two Motors

DC-DC

LM2596S DC-DC

Battery

Turnigy 2200mah 3s 20-30c 11.1v

H Bridge L298N (pinout)

5v + ENA : board jumpers enable
IN1  : 3
IN2  : 4
5V + ENB : board jumpers enable
IN3  : 5
IN4  : 6
R1   : board jumpers enable
R2   : board jumpers enable
R3   : board jumpers enable
R4   : board jumpers enabl
VCC  : LM2596S DC-DC
5V   : Intel Edison board.
GND  : Intel Edison Board and LM2596S DC-DC</p>

Simples program .

simples server UDP, commands for movimentar robot.

GiTHub : Myo_server.py

my laptop program, commands for robot.

GitHub : SenderCommands

Data Transmission Function To The Myo

In this project we used a macbook with the operating system OSX El Capitan to send the data to the Intel Edison through UDP messaging. For this, we downloaded the SDK Myo in https://developer.thalmic.com/downloadshttps://developer.thalmic.com/downloads and compile the hello-

myo.xcodeproj project within thesamples folder of the SDK.

In this example we modify the code hello-myo.cpp file in the print () function of the DataCollectorclass. In this role we have access to the name of the activated gesture at that moment in Myo. With this data in hand we formatted the UDP messages and sent over the network.
The file add the following code beginning with the declaration of variables and inclusion of libraries:

#include <arpa/inet.h>
#include <sys/socket.h>
#include <unistd.h>
// here goes the ip address of the Intel Ediso
#define SERVER "192.168.1.101"
#define BUFLEN 1024
/// here goes the port that are you using
#define PORT 21567</p><p>void die(char *s)
{
    perror(s);
    exit(1);
}

struct sockaddr_in si_other;
int s, i, slen=sizeof(si_other);
char buf[BUFLEN];
char message[BUFLEN];

After that we add the code to discover the gesture detected, format and send the message to the Python code running in the Intel Edison.

if(poseString == "fist")
{

    std::string action = "STOP_MOVE";
    std::cout << action << "; " << lastCommand;
    lastCommand = action;

   if (sendto(s, action.c_str(), strlen(action.c_str()) , 0 , (struct sockaddr *) &si_other, slen)==-1)
    {
        die("sendto()");
    }
}
else if(poseString == "fingersSpread")
{
    std::string action = "MOVE_FORWARD";
    std::cout << action << "; " << lastCommand;
    lastCommand = action;</p><p>    if (sendto(s, action.c_str(), strlen(action.c_str()) , 0 , (struct sockaddr *) &si_other, slen)==-1)
    {
        die("sendto()");
    }
}
// there is more code in the .cpp file</p>

Just simply replace the code in print() function by the code in MyoSdkSenderFunction.cpp file.

GITHUB : Intel-IoT-RoadShow-2015

Running node server.js

and myo_server.py

after used Gesture control armband.