Strategy OVERVIEW
- Success definition: Exit safe space → shoot 1 → shoot 2 → fact check
- Work on accuracy and precision rather than complexity.
- Prototype as much/fast as possible in order to iterate over successful designs and converge to a final functional solution.
- Managers are accountable for design and final implementation of their sections.
- Subleads help the leads with checking work, delegated tasks, and debugging.
Final Design
- Drive Train: 4 holonomic (omni) wheels powered by DC motors
- Shooter: Fly wheel using DC motor
- Ball Feeding: Revolver mechanism rotated by servos
- Sensors: 4 Ultrasonic sensors to determine location within the field
- Power: 2 NiMh batteries in series
- Material: Duron and foam core board
- Strategy:
- Rotate until shortest distance detected by "back" ultrasonic sensor
- Back up into safe space "back" wall
- Follow the walls until positioned directly in front of tower for consistency
- Shoot 3 balls
- Follow the wall until positioned directly in front of 2nd tower
- Shoot 3 balls
- Back up to fact check still facing the 2nd tower
- Go back to safe space to reload
- Repeat 2-7
Sensors: Ultrasonics
Mobile Platform: 3D printed parts and highly supported motors and bearings
Ball Feeder: Revolver
SHooter: ramp
SHooter: Wheel
Final DESIGN: implementation
Second Design
- Drive Train: 4 holonomic (omni) wheels powered by DC motors
- Shooter: 3D printed fly wheel using DC motor
- Ball Feeding: Revolver mechanism rotated by servos
- Sensors: 4 Ultrasonic sensors to determine location within the field
- Power: 2 NiMh batteries in series
- Material: Duron and foam core board
- Strategy:
- Rotate until shortest distance detected by "back" ultrasonic sensor
- Back up into safe space "back" wall
- Follow the walls until positioned directly in front of tower for consistency
- Shoot 3 balls
- Follow the wall until positioned directly in front of 2nd tower
- Shoot 3 balls
- Back up to fact check still facing the 2nd tower
- Go back to safe space to reload
- Repeat 2-7
Second design: Prototypes
The drive train prototype tested the use of our 3D printed parts, motors, current driving the motors, wheel alignment, and sensor algorithm. We tested 3D printed pillow blows with collars. We found that 3D printed parts allowed for a lot of flexibility with ensuring proper coupling and alignment.
Materials; duron, M3 screws/washers/nuts, spider couplers, tape, 3.25" omni wheels, jumper wires, HC-SR04 ultrasonic sensors, tape, Jameco 161382, and LM298N board.
Design change: We discovered that thinner pillow blocks needed and more support for motors needed to ensure wheel alignment. We also learned that we need to use bearings to ensure there isn't too much vibration in the shaft. Moreover, we learned that 3D printed shaft couplers and collars gave us the flexibility we needed to have proper wheel alignment.
Materials; duron, M3 screws/washers/nuts, spider couplers, tape, 3.25" omni wheels, jumper wires, HC-SR04 ultrasonic sensors, tape, Jameco 161382, and LM298N board.
Design change: We discovered that thinner pillow blocks needed and more support for motors needed to ensure wheel alignment. We also learned that we need to use bearings to ensure there isn't too much vibration in the shaft. Moreover, we learned that 3D printed shaft couplers and collars gave us the flexibility we needed to have proper wheel alignment.
The shooter prototype proved that the fly wheel was more consistent than the spring design. The most important test was to demonstrate that the 3D printed wheel, including shaft and collar, would not break during operation. We discovered that the 3D printed shaft felt too much friction in the duron and we decided to add a bearing to design.
We also wanted to determine the ramp angle and the support system needed to have consistent motor angular velocity. The tests demonstrated that the batteries provided the required power to hit the target for the given wheel radius. The length of the ramp allowed us to change the angle of launch, which proved that angles close to 0 degrees were more consistent. The other test we performed defined the appropriate location of the shooter within the robot through the use of a gantry-like support of wheel. This gave us the flexibility to change the height of wheel with respect to the ramp floor and the point of contact with the ball. We discovered that having the wheel placed 1.75" above the ramp floor allowed for best range. We also discovered that the friction tape allowed for more traction, which led to more consistent shooting.
Materials; duron, 3D printed wheel, friction tape, foam core ramp (adjustable length and angle), bearing, DC motor (Jameco 232040), shaft coupler.
Design changes: Bearing needed to ensure wheel would not shear. Ramp should have almost no angle. Friction tape helps with securing position of ball. Wheel should be 1.75" elevated from ramp floor.
We also wanted to determine the ramp angle and the support system needed to have consistent motor angular velocity. The tests demonstrated that the batteries provided the required power to hit the target for the given wheel radius. The length of the ramp allowed us to change the angle of launch, which proved that angles close to 0 degrees were more consistent. The other test we performed defined the appropriate location of the shooter within the robot through the use of a gantry-like support of wheel. This gave us the flexibility to change the height of wheel with respect to the ramp floor and the point of contact with the ball. We discovered that having the wheel placed 1.75" above the ramp floor allowed for best range. We also discovered that the friction tape allowed for more traction, which led to more consistent shooting.
Materials; duron, 3D printed wheel, friction tape, foam core ramp (adjustable length and angle), bearing, DC motor (Jameco 232040), shaft coupler.
Design changes: Bearing needed to ensure wheel would not shear. Ramp should have almost no angle. Friction tape helps with securing position of ball. Wheel should be 1.75" elevated from ramp floor.
Original Design
Our original design did not persist through the prototyping phase as we had to make changes in order to ensure proper functionality.
- Drive Train: 4 holonomic (omni) wheels powered by DC motors
- Shooter: Spring loaded with use of steel flat plate displaced by servos arm
- Ball Feeding: Servos arm regulator to serve as containment area with use of ramps
- Sensors: 2 Ultrasonic sensors to determine nearest wall in safe space and tape sensors to navigate
- Power: 2 NiMH batteries in series
- Material: Duron
- Strategy:
- Rotate until shortest distance detected by "back" ultrasonic sensor
- Back up into safe space "back" wall
- Follow the walls until positioned directly in front of tower for consistency
- Shoot 6 balls
- Follow the wall until positioned directly in front of 2nd tower
- Shoot 6 balls
- Back up to fact check still facing the 2nd tower
Original design: Prototypes
This ball feeder prototype showed us that with our limited vertical space, this design would not be optimal in order to fit 12 balls and have balls slide consistently. Moreover, we discovered that balls would get easily stuck between levels and therefore opted to change our design for a revolver mechanism.
Materials: Foam core board, duct tape, and clamps
Design change: from slide to revolver
Materials: Foam core board, duct tape, and clamps
Design change: from slide to revolver
The shooter prototype showed us that the spring mechanism was not consistent because the work done by the steel plate and the angle of launch depended on the placement of the ball. Hence, we discovered that feeding balls to this design was problematic and required high levels of accuracy that we could not get with the ball feeder design. The steel flat plate required a servos much more powerful than the one we had calculated: 3kg-cm (HS-322HD). We looked into purchasing others at Fry's with 13 kg-cm (LS8101F) but our budget did not allow for such an expense. Therefore, we moved to the fly wheel design.
We made the prototype to figure out the appropriate length and angle of the steel plate. The clamps allowed for the adjusting of length and the brackets used to clamp down the plate could be rotated to find the appropriate angle needed to launch the ball to appropriate distances.
Materials; Birch wood, poplar dowels, steel flat plate, HS-322HD RC Servos, duron servos arm, clamps
Design change: from spring to fly wheel
We made the prototype to figure out the appropriate length and angle of the steel plate. The clamps allowed for the adjusting of length and the brackets used to clamp down the plate could be rotated to find the appropriate angle needed to launch the ball to appropriate distances.
Materials; Birch wood, poplar dowels, steel flat plate, HS-322HD RC Servos, duron servos arm, clamps
Design change: from spring to fly wheel
The drive train prototype showed us that the DC motors ( Jameco 161382) and wheels could drive us in a straight path with a given weight. The main goal of the prototype was to determine wether the ultrasonic sensors would be able to function properly without too much reflection at the elevation of the drive train. Our tests concluded that the designed positioning worked well with the software logic.
Materials; duron, M3 screws/washers/nuts, spider couplers, tape, 3.25" omni wheels, jumper wires, HC-SR04 ultrasonic sensors, tape, Jameco 161382, and LM298N board.
Design change: Addition of 3D printed pillow blocks and using bearings. Use of 4 ultrasonic sensors instead of only 2 because our testing demonstrated that we got proper readings from all 4 walls for our given algorithm.
Materials; duron, M3 screws/washers/nuts, spider couplers, tape, 3.25" omni wheels, jumper wires, HC-SR04 ultrasonic sensors, tape, Jameco 161382, and LM298N board.
Design change: Addition of 3D printed pillow blocks and using bearings. Use of 4 ultrasonic sensors instead of only 2 because our testing demonstrated that we got proper readings from all 4 walls for our given algorithm.