Society of Robots
Search and Index Search Here

 Parts List
 Robot Forum
 Member Pages
 Axon MCU
 Robot Books

 How To Build
  A Robot




 Robot Journals
 Robot Theory


    The Original Taurus
    Taurus was my very first robot, named so as it was planned to follow anything the color red. But being my first robot, built in early 2002 as a sophomore, it was an absolute failure. And so a year later, in late 2003, I decided to redo it - knowing everything I learned from building several other robots in between. Also, you may note that the wheel kinematics are similar to the Mars rovers. I must emphasize - I built this robot before ever being aware of the rover design!

    Taurus 2

    Taurus 2 Mechanical Features
    Has a zero degree turning radius (can rotate in place). It has near omni-directional movement, minus singularities at maximum servo rotations and the time it takes to turn a wheel. This allows Taurus 2 to travel sideways, forwards, straff, and even drive in sinusoidal shapes - yet still face in the same direction. And lastly, it can handle rough terrain.

    Here is a composite video of T2:

    ( music by Zox - Canon )

    Mechanical Design
    T2 was made in 2D CAD (my first time using CAD) and built using HDPE and aluminum as the chassis material. A long metal rod attached to two ball bearings created the rocker-bogey system, as shown in this assembly picture:

    T2 Rocker-Bogey Design

    Electrical Design
    I used a PIC based microcontroller programmed in C. Eight digital outputs controlled each of the servos independently, while the CMUcam was interfaced by the serial port. The eight servos drained large amounts of power, so I had to use two NiCad battery packs. As additional contactless 'bumper' sensors I made three infrared emitter detector sensors, as shown here:

    Infrared Emitter Detector Pair

    Programming and Control
    I programmed T2 to read in the middle mass (centroid of a color in an image) of the colors green and blue in whatever it saw from the camera. I then would have the servos rotate in such a way so that T2 would always drive to that middle mass. I wanted T2 to follow a white line, and so this worked really well. To control four wheels independently, you must have each wheel turn to a specific angle so that the axis of each wheel goes to a single central point (or axis of rotation). This image will explain it well:

    Four Wheel Angle Kinematics

    If the front two wheels had equal angles, the rotation axis will not be common, and you will have high friction and skidding. If the wheels were to never change the angles as shown in the above image, the robot would follow the imaginary circle perfectly about the common central axis. Just use basic trigonometry, knowing the turn radius and wheel locations, to determine required wheel angles.

    When I programmed it, I made a continuous fuzzy logic algorithm to relate wheel angles to target direction. The front wheels and back wheels were inverted so I just needed to do two angle calculations per cycle. I also did a few mixes of other algorithms, as seen in the above video. Another method is to make a trigonometric lookup table to control wheel angles.

    This is the wheels positioned so that T2 can rotate in place:

    Four Wheels, Angled

    And one more closeup image:

    Taurus 2 Closeup

    Parts Used
    The total cost to me to build the entire robot was about ~$90, with many parts being scrapped from a few of my older robots. Here is a list of parts and typical prices of everything used:

    (2) 7.2V nicad battery packs ($16)
    (1) CMUcam ($100)
    (1) Cerebellum Microcontroller, by Botrics ($40)
    (3) homemade IR (~$4 in components)
    (8) HS-311 servos ($71)
    (1) null modem cable ($3)
    (4) foam 4 inch diameter model aircraft wheels ($15)
    (1) 1'x1' sheet of HDPE ($10)
    (2) ball bearings (found in a scrap pile)
    (1) 1'x1' sheet of aluminum ($10)

    Make Money by Making Robots
    Building Taurus didnt actual cost me anything, as not only was the cost paid for by a grant that I applied to (with extra $$ left over), I also entered this robot into the 2004 CMU MOBOT competition - winning 3rd place and $250.

    MOBOT check

    Dont Use the CMUcam!
    I grumble that I didnt place 2nd that year. Apparently the CMUcam has a not very well known bug, causing the firmware to crash when getting large amounts of blue light - such as from the sun. I couldnt figure out why it would just stop working, until 2 years later when looking at the CMUcam output in sunlight in detail. The developers have never fixed this problem and have no plans to. Dont believe me? Point it at the sun during a sunny day and watch it stop working! (rant rant grumble grumble)

    Source Code:
    PIC 16F877 source code for Taurus2
    note: I can't seem to find my most updated code, so this is take it as it is.

Get Your Ad Here

Has this site helped you with your robot? Give us credit - link back, and help others in the forums!
Society of Robots copyright 2005-2014
forum SMF post simple machines