1. Stanley – RC Car

2. Introduction The idea of this project Initial Idea Add on
develop Remote Controlled Car that will be controlled by the computer.
Initial Idea
Control Stanley by the keyboard
Control Stanley by the mouse
Draw the path
Add on
Direction by laser pointer
Follow the Light point
Follow the object based on the colour.

3. The Hardware Remote Controlled Car PhidgetInterfaceKit 0/0/4 board
Digital controller
PhidgetInterfaceKit 0/0/4 board
USB-based controller,
with 4 digital outputs
USB Webcam
IP Webcam

4. Connecting the Car Controller Identify the points
Switches for controlling
directions
One switch = two points
(ground and active)
Press switch closes circuit
Identify the points
Different colour wires for different direction
Solder on group of two points

5. Connecting the Car PhidgetInterfaceKit Board Screw the wires
4 Relay Outputs, each
relay has 3 connections:
#C = Common
NO – Normally Open
NC – Normally Close
Screw the wires
Grouped by the colour – to each relay
#C, and NO
Port no = #

6. Connecting the Camera Power connection wireless
Power connector (2mm out diameter
1.3 inner diameter).
Battery Holder
Wires
4 AA CHARGABLE batteries at 1.2V each, with capacity rating over 2000mAh.

7. The Software Visual Studio 2008 Microsoft Robotics Studio
Phidget Libraries
Aforge Libraries

8. The Software Microsoft Robotics Studio
Microsoft Robotics Studio Runtime have two main components;
Concurrency and Coordinate Runtime (CCR)
Managed code Library
The Decentralized System Services (DSS)
Decentralized Software Service Protocol (DSSP) –to interact with the services
XML/Soap
Subscribe to service
Contains state information, identification and has ports that can accept messages and service handlers to process these messages
Supports wide range of programming Languages; VB.NET, C#, Microsoft Visual Programming Language (VPL) (provided as part of the Microsoft Robotics Studio)
Create Service
Command prompt
dssnewservice/service: RemoteCar
Folder with project and Solution Files
RemoteCar.cs (contains implementation of the remote car service)
RemoteCar.manifest.xml (indicate a set of service that are to be started)

9. The Software Phidget Libraries
PhitgetInterfaceKit is managed behind this library
MRS includes a built-in service for PhidgetInterfaceKit
Communicate with the PhidgetInterface
Add Reference
Import Library
private phidgetinterfacekit.PhidgetInterfaceKitBoardsOperations _ikPort = new phidgetinterfacekit.PhidgetInterfaceKitBoardsOperations();
Declare Port
private phidgetinterfacekit.PhidgetInterfaceKitBoardsOperations _ikPort = new phidgetinterfacekit.PhidgetInterfaceKitBoardsOperations();
Subscribe To the Service
_ikPort.SelectiveSubscribe(null,new phidgetinterfacekit.PhidgetInterfaceKitBoardsOperations());

10. The Software public enum Direction { Forward = 0, Backward, Left,
Right }
public void SetOutput(Direction dir, bool enabled)
if (dir == Direction.None)
return;
// create a new request to send to the Phidgets service
phidgetinterfacekit.SetOutputRequestType req = new phidgetinterfacekit.SetOutputRequestType();
// assign the index and state (true/false)
req.Index = (int)dir;
req.State = enabled;
// send the request to the port
_ikPort.SetOutput(req);

11. The Software Aforge.NET Libraries Developed by Andrew Kirillow
Written in C#
For Computer Vision
6 main libraries:
AForge.Imaging is library for processing routines and filters.
AForge.Neuro is neural networks computation library.
AForge.Genetic is evaluation programming library.
AForge.Vision is computer vision library.
AForge.Machine Learning is machine learning library.
AForge.Robotics.Lego is for manipulation of Lego Mindstorm NXT devices.
Classes for accessing local video capture devices (DirectShow), and motion JPEG stream.

12. Stanley without camera
Control the car by keyboard
Testing tool
ProcessCmdKey/keyUp
Control the car on the screen
MouseUp/Down
Control the car by the mouse
Compare last x and y coordinates with new (count every 5th)

13. Stanley without camera
Control the car by the mouse
Compare last x and y coordinates with new (count every 5th)

14. Stanley without camera
Follow the Path
Compare first x and y points with last
Calculate line length between points
(Line Length = √ ((x2-x1)*(x2-x1)) + ((y2-y1)*(y2-y1))
Calculate Angle
Compare angles ==>
Measure Stanley
movements

15. Stanley with camera Control on the screen With camera
Testing tool
MouseDown/Up
Follow direction from the laser pointer
Look trough each pixel in image
Convert each pixel in corresponding colour from grey scale by using the Greyscale BT709 algorithm
red * green * blue * = 0-255
Value >240
Record x and y points
Determine direction
Same algorithm like in “mouse move”

16. Stanley with camera Follow the Lightest point
Find the lightest pixel – same as in previous example
Record the x and y points
Compare the points with camera position
Determine direction
Determine Distance
(stop/go)
Search for the object

17. Stanley with camera Track the object based on the colour Filter Image:
Blur
Colour Filter –RGB – range 0-255
Convert to Greyscale
Threshold
Convert image into a 1-bit binary image
Black and white (depending on the pixel value
Get Image size and position
BlobCounter / getObjectRectangles();
Find the centre of the object
Record the x and y points
Compare the points with
camera position
Determine direction
Determine Distance (stop/go)
Search for the object

18. Deploy the Stanley DSS Deploy Tool
Creates files containing all MRS components needed for running the application
dssdeploy /p /m:“RemoteCar\RemoteCar.manifest.xml" InstallStanley.exe
Start
MRS not needed
Trough Standard Command Prompt
cd /d C:/InstallStanley" bin\dsshost /p:50000 /m:" RemoteCar\RemoteCar.manifest.xml"