The Cone-Finder


The Robot

The picture above is of a highly sophisticated robot designed to do one thing: find orange traffic cones.  This enormous undertaking has consumed the last nine months in getting it to this stage, and will take another six to complete for entry in the Robo-Magellan contest.  The Robo-Magellan is part of a group of contests called the Robothon, which is held annually in Seattle by the Seattle Robotics Society.  Robo-Magellan contestants design and build a robot that can find orange traffic cones on an outdoor course.  This robot will compete in this event, finding orange traffic cones and avoiding obstacles.  The robot will negotiate these obstacles through the use of specialized sensors.  These sensors consist of:

  • A GPS to locate the robot in position to the traffic cones that have GPS coordinates associated with them
  • A one-dimensional camera to detect the overhanging obstructions, such as trees, that might disrupt GPS signals
  • A compass that backs up the GPS in situations where the GPS does not perform well
  • Distance sensors to detect obstructions in front of the robot
  • An angle sensor to sense hills for safety
  • A color-tracking camera to find and track the orange traffic cones while the robot is moving

Together, these sensors will help guide the robot to victory!

The Contest

The Robo-Magellan is hosted in Seattle every year by the Seattle Robotics Society as part of their annual Robothon.  The contest is staged outdoors where there are orange traffic cones scattered around.  In the contest there are a specific number of cones that have GPS coordinates associated with them.  The robot must touch each cone once to complete the course.  These cones are strategically placedunder trees and other overhangs to distort GPS.  Each team gets three runs each lasting 15 minutes, with 30 minutes in between to make software and hardware adjustments.

The Brain

The “brain” of the robot consists of a computer chip, or microcontroller, and a circuit board.  The particular microcontroller that the robot is using is called an Arduino Mega and has 54 data pins where sensors can be connected.  One aspect of the Arduino that is unique from most other microcontrollers is that it can be connected by USB to the computer with the Arduino software, instead of a classic serial port, which is an advantage because most modern computers do not have a serial port.

The Sensors

SensorGPS, A Future Addition to the Robot

Use on the Robot

Before the contest the GPS coordinates of the traffic cones are given to the contestants.  The GPS will help locate the robot in position to the traffic cones.  At the beginning of the contest the microcontroller will ask the GPS for the current coordinates of the robot.  The microcontroller will then compute these coordinates with the coordinates of the first traffic cone, coming up in the end with the speed, direction, and distance the robot would need to go to reach the cone.

:   TSL-1401-DB, Line Scan Module

Use on the Robot

The line scan module will help to detect if the robot is under a tree, as there will be trees in the contest.  Being under a tree is detrimental because there is no GPS signal there.  But one problem with using GPS in the Robo-Megellan contest is that the cones are strategically placed under trees or other things that might block a GPS’s signal.  One remedy for this is to use GPS to direct the robot to the cone for as long as is possible, or in other words, until the line scanner senses a tree limb.  After the line scanner senses a tree limb, the microcontroller will stop using the data from the GPS and start using some of the other sensers to compensate for the loss.

What it Does

The line scanner consists of a single row of 128 photodetectors. The line scan module also includes a lens to form images on the sensor array.  This is like looking through a crack of a partially open door to see a thin slice of what lies behind it.  An example of this is shown at the left.



SensorHMC-6352, Compass

Use on the Robot

The compass sensor is an electronic compass whose job on the robot is to take over where the GPS fails, such as under a tree and will also act as a check for the accuracy of the GPS.

How it Works

The compass sensor is a 2-axis magneto-resistive compass.  A magneto resistive material is a material, discovered in 1856 by Lord Kelvin, which has the property to change its electrical resistance when a magnetic field is applied to it.

SensorPing, Ultrasonic Range Finders

Use on the Robot

The ultrasonic range finders mounted on the robot will detect whether an object, such as a cone, a bush, or a tree is in the path of the robot.  There is a sensor on each side of the robot that detects objects to either side of the robot.  To detect whether an obstacle is directly in front of the robot, all that needs to be done is to compare the two sensor readings, as shown at the right.

What it Does

There are numerous ultrasonic range finders on the market, but most have more than one data pin.  Having only one data pin is superior because it makes coding much easier.  An ultrasonic range finder works by sending out little chirps at 40 kHz.  These chirps echo off of an object and are received by the sensor.  The time it takes from the time the range finder sends out the chirp to the time it is received is converted to distance in the microcontroller.


SensorIXZ-500, 2 Axis Gyroscope

Use on the Robot

The gyroscope will be used to determine whether the robot is on a hill and which way it is facing on that hill.  A gyroscope is a device for measuring or maintaining orientation.  For example, if the robot knows that the first cone in the contest is on the top of a hill as well as the grade of that hill, the robot will then infer from that using the information from the gyroscope whether it is getting close to the cone.

SensorCMU Camera

Use on the Robot

The CMU camera is the most expensive single part on the robot as well as the most important.  The CMU camera’s job on the robot is to catch sight of the traffic cones as well as track them while the robot is moving.

What it Does

CMU doesn’t stand for anything technical, it just stands for Carnegie Mellon University because it was developed by their robotics institute.  The camera is specifically designed for robots that need to find and track colorful blobs, such as traffic cones.  The CMU consists of a regular camera in front that sends the video signal through a special computer chip that, in turn, sends the coordinates of what it is seeing to the microcontroller.


Special Thanks

Special thanks to Jim for his patience, guidance, and support.

3 Responses to The Cone-Finder

  1. hatem elborai says:

    i want to buy a kit

  2. rajeev says:

    Nice project. i am also working on a similar project with cmucam but on a smaller scale. could you please provide me the codes for interacting arduino to cmucam and how to control the respective motors to avoid obstacles.

    This will be a great assistance my frend.Thanks

    • diggerdata says:

      Sorry to disappoint you, but I have not written the code you are requesting and am not intending to because I am going to use and eee PC with computer vision software instead. Maybe some day I will write an Arduino library for interfacing with the CMU cam.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: