Theme
This year's theme was to launch frisbee's on top of those poles shown above, displacing the balls (or the opponent's frisbees!) and getting as many frisbees as possible before the time ran out. The contender with more points (more frisbees) was the winner. I was responsible for the mechanical design of the robot.
How we did it
We first experimented with various launchers - pneumatic and electric powered. For the frisbees provided to us by the organizers, we finally settled on a mechanism where the frisbee would be pushed between a smooth surface and a high-speed wheel with friction padding on it, similar to the one shown on the right.
I then settled on the degrees of freedom needed by the launcher to accurately launch the frisbees to its target. We provided pitch to aim for the varying heights of platforms to land the frisbees on; and we used a rack-and-pinion to do this. Further, it would be advantageous to minimize the amount of movement of the mobile robot to save time and launch from a single pose, or at least provide some fine yaw motion to aim across the arena.
The requirements were that the yaw mechanism was both precise and able to carry the load of the all the mechanisms - the launcher, the frisbee feeder and the pitch mechanism. I implemented a planetary gear mechanism it as it satisfied all the requirements - high-resolution motion, handling of heavy loads without high power requirements from the driving motors (this was as opposed to a single motor in the center that would require huge amounts of torque to rotate the same load).
However, in the competition we faced one setback, we practiced with frisbees that worked well with high spinning motion and didn't require the roll motion (the roll motion is used to provide the "angle of attack" which counters the tilt experienced by the robot due to the fact that the aerodynamic forces acting on the surface with which the air particles moving relative to the frisbee collide).
The frisbee which we actually received in the competition (changed by the organizers at the 11th hour) were more rigid and in contrast with the soft frisbees which would have an elastic collision with the surface and not rebound significantly; the new frisbees would rebound as a result of their rigidity counter to the direction of spin. Therefore, our robot was far less performant in the actual competition.
What I learnt?
- There were several constraints / requirements that needed to be met before designing this robot. But the approach I developed in the course of this experience was to first eliminate the unknowns before executing a design that meets all requirements. This was the the case for the launcher.
- The competition rules stated specific dimensional and weight requirements which had to be met, which required me to design components appropriately and use the material with the lowest density that could handle the dynamic loads. To minimize the weight of heavy bolts, I selected TIG welding to be done on to join the pieces of aluminum that would form the chassis.
- Also, selecting the right actuators meant doing the required speed and torque calculations accurately where I used worst-case scenarios to guide the requirements (For example, assume that the motors driving the wheels that move the robot around would carry the maximum allowable weight around). The reason we were able to get away with such small motors for the planetary gear mechanism was because the speed requirements for the yaw mechanism were pretty low, and the torque provided by each of the 2 motors was only equal to the force needed to "roll" the planet gears on the sun gear times the radius of the planet gears (divided by 2, for each motor).
- And of course, their were budget constraints. The manufacturing for all components had to be done by external vendors, as we didn't have access to the machine shop. So, I had to select materials, actuators and manufacturing processes that would meet the cost requirements.
Displaying the robot at our College's Annual Exhibition
Me and my "year" colleagues at the competition