Robot Design & Function
Design Challenges
The robot was designed to collect as many ping-pong balls as possible during a two minute round. While brainstorming solutions, several major design challenges quickly emerged.
- How to collect ping-pong balls off of the playing field floor
- How to add ping-pong balls to a storage container
- How to dump balls into a scoring bin
Collecting Ping-Pong Balls
The leading candidate for a ball collecting mechanism was dubbed the "Paddle Wheel." The Paddle Wheel would rotate using two identical wheels with a radius larger than the diameter of a ping-pong ball. The relatively large size of the device would allow it to encompass a ball. Several arms (paddles) would attach to the axle that connected the two wheels. These arms would sweep balls off of the playing field floor, onto a scoop, and then onto a platform within the robot. In order to conserve power, the Paddle Wheel would not be driven by a separate motor, but would passively rotate as the robot moved forward. A prototype of this device was constructed (see Prototypes section) and after successful testing was adopted as the solution to gathering ping-pong balls.
Figure 3 shows the Paddle Wheel and Scoop on the robot. Foam core was used to construct the wheels and arms. The arms were fixed to a wooden dowel, which served as the axle for the wheels. The dowel was press fit into ball bearings so that it could rotate freely with little resistance from friction. A scoop was created from aluminum so that it could easily be fashioned into the desired shape. Additional figures of the ball collector can be seen on the Hardware page.
The Paddle Wheel function is summarized below.
The Paddle Wheel function is summarized below.
- The Paddle Wheel rotates freely as the robot moves forward.
- Stationary ping-pong balls are encompassed by the Paddle Wheel.
- The paddles sweep balls onto the aluminum Scoop and up, onto a platform, inside of the robot.
Adding Ping-Pong Balls to Storage Container
With the strategy of collecting as many ping-pong balls as possible, the robot required a method of adding newly collected balls to the storage container. Because of the gravitational forces acting on the balls, it was decided that balls should be dropped into the storage container from near its top, rather than pushed in from its bottom. Several concepts were developed and some were prototyped (see Lead Screw on the Prototypes page). After several rounds of testing, it was found that using the suction created by a computer fan was a simple yet effective means of lifting the balls into the storage container. A "Ball Funnel" was designed (see Hardware page) to constrict the area through which balls were sucked and increase the vacuum force on the ping-pong balls.
A tube (Figure 4) to contain the ping-pong balls as they were sucked from the collection platform to the storage container was formed from aluminum. The position of both the fan and tube within the Storage Container is shown in Figure 5.
The operation of the Ball Funnel, tube, and fan are summarized below.
The operation of the Ball Funnel, tube, and fan are summarized below.
- Balls on a platform inside of the robot (collected by the Paddle Wheel) are narrowed into a smaller area by the Ball Funnel.
- The fan creates a vacuum force that sucks ping-pong palls through the tube into the Storage Container.
Dumping Ping-Pong Balls into Scoring Bin
Several ideas were initially conceived for dumping ping-pong balls into a scoring bin. At first, plans involved raising up the Storage Container using lead screws, a scissor lift, or some other method; however, these solutions all required some sort of complex mechanism to achieve. It was soon realized that the Storage Container did not have to be lifted very high to clear the bin wall, which was only 3.5 inches tall. A much simpler design that placed the ball container such that its bottom was higher above ground than the bin wall, and that had a door that would open when it was time to dump was adopted. The floor of the Storage Container was slanted so that the balls would roll out into the bin once the door was opened. Originally, it was believed that a servo would be needed to actuate the door on the Storage Container, but after running several tests, it was found that a foam door would stay shut because of the vacuum forces acting on the door while the fan was running. The door would also open because of gravitational forces once the fan was switched off. Thus, the dilemma of dumping balls was solved by simply controlling the fan.
Robot Game Function
To begin the game, the robot was placed into the starting box and aimed at the beacon to its right on the opposite side of the playing field. A colored LED turned on signifying whether the robot was on the red team or the blue team. After wirelessly receiving a start command, the robot began its ball collection routine.
The robot drove straight toward the rotating wall, and turned right just before making contact with the wall. It then drove along a path parallel to the rotating wall until it reached the playing field boundary. It made a clockwise turn, reversed until a magnet located on the back of the robot attached to the rotating wall, and started driving forward on a large radius. It followed along the playing field wall, dragging the rotating wall behind it. When any of the four front limit switches (bumpers) were triggered, the robot entered a "random walk" mode. Under this mode, the robot traveled in straight lines until one of its bumpers was triggered. These bumpers were constructed of foam core strips attached to limit switches using springs and glue. Two were placed on the back corners of the robot, two were placed low on the front corners, and two were placed high on the front corners. Once a bumper was triggered in this mode, the robot stopped, reversed slightly, rotated 100 degrees clockwise, and then drove straight again. It continued in this mode for a set period of time in order to collect the balls near the middle of the field that it could not get with its large circumference sweep.
After 90 seconds passed, the robot entered the ball dumping state. The robot searched for an available bin -- determined by the angle of the rotating wall -- and drove to it, until it hit the playing field wall or the rotating wall. If it hit the rotating wall it would search for a new available bin. If it hit the playing field wall, it drove along the playing field wall until a front sensor was triggered and then reversed until it saw tape. Once tape was found, it would rotate toward the opposite bin and align itself using two tape sensors and the IR sensor. Then it would back up until the rear sensors were both triggered and turn the fan off, lowering the door and releasing the balls into the bin. This was all timed so that there would be less than ~5 seconds left in the game to decrease the likelihood of bin cancellation by the other team.
The robot drove straight toward the rotating wall, and turned right just before making contact with the wall. It then drove along a path parallel to the rotating wall until it reached the playing field boundary. It made a clockwise turn, reversed until a magnet located on the back of the robot attached to the rotating wall, and started driving forward on a large radius. It followed along the playing field wall, dragging the rotating wall behind it. When any of the four front limit switches (bumpers) were triggered, the robot entered a "random walk" mode. Under this mode, the robot traveled in straight lines until one of its bumpers was triggered. These bumpers were constructed of foam core strips attached to limit switches using springs and glue. Two were placed on the back corners of the robot, two were placed low on the front corners, and two were placed high on the front corners. Once a bumper was triggered in this mode, the robot stopped, reversed slightly, rotated 100 degrees clockwise, and then drove straight again. It continued in this mode for a set period of time in order to collect the balls near the middle of the field that it could not get with its large circumference sweep.
After 90 seconds passed, the robot entered the ball dumping state. The robot searched for an available bin -- determined by the angle of the rotating wall -- and drove to it, until it hit the playing field wall or the rotating wall. If it hit the rotating wall it would search for a new available bin. If it hit the playing field wall, it drove along the playing field wall until a front sensor was triggered and then reversed until it saw tape. Once tape was found, it would rotate toward the opposite bin and align itself using two tape sensors and the IR sensor. Then it would back up until the rear sensors were both triggered and turn the fan off, lowering the door and releasing the balls into the bin. This was all timed so that there would be less than ~5 seconds left in the game to decrease the likelihood of bin cancellation by the other team.