The First for Inspiration and Recognition of Science and Technology (First) Robotics Competition immerses high-school students in the engineering realities of meeting project objectives while coping with trade-offs, timelines and constraints. Teams of 10 to 20 students, plus their engineer mentors, are each given a standard kit of parts with which to complete a common design challenge within six weeks. The following account of one team's experiences in the 2007 and 2008 competitions was written by a student participant.
With only six weeks to design and build a robot, teams that compete in the First Robotics Competition must be adept at effective, efficient design and ready to tackle any challenges that come their way. Team 418--aka Purple Haze--from the Liberal Arts and Science Academy (LASA) High School in Austin, Texas, has competed for the past eight years. The group faced particularly knotty design challenges in 2007 and 2008.
When designing this year's Zephyr robot, Team 418 kept an eye on two factors: ergonomics and autonomy. The team needed to devise a simple, easily removable bumper system, and got one by implementing a bracket that slid into the aluminum frame of the chassis and attached using screw eyes. The ergonomic design made the robot easy to lift, and Zephyr worked without a hitch in autonomous mode.
For the prior year's meet, the LASA team faced two primary challenges: designing a suitable and rugged drive train that could withstand the rigors of the competition, and developing an intuitive control mechanism for the robot arm that would let operator with little experience use the arm flawlessly.
For the 2007 "Rack 'in' Roll" competition, robots had to place inner tubes on moving octagonal racks while fending off the blocking attempts of their opponents. LASA's team built PHunky, a six-wheeled robot that manipulated the game pieces (inflatable pool tubes) with a double-articulated arm. Coming up with the right drive train was difficult, because both power and speed were important in different aspects of the game: power to defend and score, and speed to "reload" the robot manipulator and move it across the large game field to garner the perfect position.
Finding inspiration in a white paper written by other First teams, the LASA team began to develop and refine its own version of "DeWalts," devices that consisted of First-provided CIM motors and DeWalt XRP drill transmissions. The ability of these transmissions to shift among three gears let the team switch reduction ratios on the fly.
To push obstacles out of the way, PHunky's 12:1 I/O ratio in first gear provided more than enough power to overcome whatever was in its path. When in second gear, the 120-pound machine could zip along at 14 feet/second, speedy enough to outrun the competition. The mechanical team mated kit-supplied motors to the drill transmissions via custom mounting plates created on a two-axis CNC mill. Shifting was accomplished by connecting servos and solid copper wires to the speed-selection lever on the transmission.
Although no CIM motors had ever failed in the team's experience at First, team members knew the DeWalt assemblies alone were not enough to ensure PHunky wouldn't become a sitting duck in the extreme case of motor failure. With two motors powering each side of the robot, the team needed redundancy so that should one motor fail, its respective side could continue to run off the other motor's output.
With this in mind, the team decided on a system in which two sprockets on the central wheel's axle connected to the front and rear axles via a No. 25 chain inside the chassis, while the two motors on each side powered the center and rear wheels on the chassis' exterior.
The controls team, for its part, focused its effort on the business end of the robot: the double-articulated arm and manipulator. During development and initial testing, the controls team used simple joysticks to regulate the movement of the arm, with one joystick controlling each articulation. This provided precise control over the arm, but the robot operator found it very clunky to use. While practicing, the operator accidentally damaged the arm by running it into the ground and against the rack built by the team in its shop.
The team quickly developed a new control system for the robot operator, dubbing it the Lewis Arm after the student who came up with the idea. The Lewis Arm was a one-sixth scale model of the actual double-articulated robot arm, with potentiometers mounted on the articulations. (Potentiometers were used instead of digital encoders because of hardware limitations imposed by the competition-mandated control system.)
Team members mounted similar potentiometers on the actual robot articulations and calibrated the potentiometers on both systems so that a given position on the Lewis Arm matched with a known possible position on the actual robot. That let the operator manipulate the robot in a mimic fashion, "posing" the control arm to get the robot arm to move into the same position. The team placed mechanical stops on the Lewis Arm to prevent it from allowing the actual arm to meet any harm. Using the Lewis Arm was much more intuitive than using joysticks.
The control algorithm for the arm was a bit more complicated to write and test. With four days left before the ship date, and with the robot arm mounted to the robot, team members could not rigorously test the arm control algorithms safely on the robot itself, which was constantly being tweaked during all hours of the day. Because testing forced all work to cease, it couldn't be done for long periods without hampering the team's ability to finish the robot on time.
 |
|
Team Purple Haze poses with its First competition entry.
Inset: The Lewis Arm control mechanism, with potentiometers on the articulated joints |
