The Frisbee Flinger

About

The Frisbee Flinger is an automated frisbee thrower designed to simulate various passes in Ultimate Frisbee and allow players to practice without a partner.

Team Members

Lindsey Lu

Future Ultimate player, sunset enthusiast, most likely to get stuff done early

Rory McDowell

Avid frisbee catcher, mediocre Ultimate player, and likely Mechanical Engineering major

Jack Flaggert

Ultimate frisbee player, professional water gun safety instructor

Sam Aronson

Amateur wrangler, Ultimate player, and Electrical Engineering major, Most likely to get stuff done on time (barely)

Problem Statement

Currently, competitive frisbee players have no consistent and reliable way to practice catching alone because no automated frisbee thrower exists.

The Scope:

User & Purchaser

Competitive Ultimate Frisbee Teams:

State of the Art

Automatic Flying Disc Launcher

Designed for playing with dogs: it lacks angle control and does not fit the standard Ultimate disc.
Patent: US 11015896 B1

Mechanical “Arm” Disc Launcher

This is a fully handheld, mechanical device with the launch speed, power, and rotation dependent on human control.
Patent US 7900617 B1

Homemade Projects

Homemade projects, such as one by Mark Rober (above) involve a spinning that launch the disc in a circular motion. However, there is no control over speed or rotational motion.

Specifications

With humans as our control, our goal is to create a launcher that is better and more consistent than practicing with a partner.

Design and Prototyping Process

Foam-core models

1:1 foam core model to layout design

CAD designs

CAD drawings were created to finalize and quantify designs

Design Drawings

Initial drawings showed roughly how pieces would attach

Early Models

3D printing models provided a more precise prototype than foam


Initial concept design (CAD using TinkerCAD)

The Final Prototype

Final prototype features an independent motion system. First, a wheel spins up the frisbee which is held in place with bearings. Next, an air cylinder accelerates the frame and frisbee to the appropriate launch speed, allowing the frisbee to be ejected from the front

Benchmarking & Testing

Human Spin Benchmarking

On a short, slow throw, a disc rotates at an average 6.24 revolutions per second (or ~375 rpm)

Launcher Spin Performance

In the ideal form (position with less friction), a disc reaches around 600 rpm, which is much greater than a short human throw and mimics the rotational speed necessary for a long distance throw.

Human Accuracy Benchmarking

Rory (an Ultimate player) and Lindsey (not an Ultimate player) tested accuracy on a shotput field with a target for the star. The green circles represent positions of where the disc landed. Though Rory was more consistent than Lindsey, both were relatively inaccurate.

Launcher Accuracy Performance

At the launcher's current range, the disc consistently hits the same spot. With further improvements to increase distance, the launcher is expected to maintain consistency.

Specification Analysis

Consistent Spin:
Spin was consistently at 67 rpm in current form, and 600 rpm in ideal form, reaching specifications

Controllable Release Angle:
Angle control was not built into the device due to time constraints, however, the user could tilt the machine to create angled throws

Controllable Distance:
Distance was consistently 5 feet, however, range was limited due to PSI available and mechanical recoil

Cost:
Cost was around $560 in materials, however, mass production would drop this below the target of $500

Transportable:
The device 3.5ft x 1.5ft x 1.5ft, is well within the specifications of <5 cubic ft.

Magazine of Discs:
The device currently only stores one disc due to time constraints

Damage to Discs:
Discs showed no additional damage after testing

Safe:
Once a cover is added, the device will become no more dangerous than a human throwing an object. At this stage in development, there is still a risk of injury from the moving parts

Automatable:
The current device requires two users, one to catch and one to facilitate the launch, but full automation of the launch is possible for future implementation.

User Feedback

We reached out to the frisbee teams of Dartmouth College by sending a survey asking for feedback on the current prototype. All respondents were shown videos of the prototype firing from two angles, then asked to rate various attributes of the prototype on a scale from 0 (worst) to 10 (best). The results of the survey are shown in the graphs below.

Spin

Respondents rated the spin on the throws from the prototype 

Release Angles

Respondents rated the range of release angles of throws from the prototype

Distance

Respondents rated the distance of the throws from the prototype

Portability

Respondents rated the distance of the throws from the prototype

The feedback on our prototype was neither seriously positive or negative, with most responses in the middle ground. Respondents were neutral on the portability (average rating of 4.4) and release angles (average rating of 4.6), but thought the spin of the prototype was good (average rating of 6.5). Unfortunately, the distance of our prototype received largely negative reviews (average rating of 1.5) which prompted us to target that area even more aggressively as a place for improvement (see our future Improvements section for more details).

Ethics and Sustainability

Highlighted materials replaced PLA for more sustainability. Manufacturing is modular to allow individual parts to be easily replaced. Additionally, a single rail would be used to cut down on metal. The wheel height would be reduced to save weight, and a fuse would be added to help prevent the motor from burning out.

Future Improvements

Automation: Our current prototype requires a human to load and fire it. We intended our prototype to be automated but ran out of time to implement it. Automation is our simplest and most important future improvement. We simply need to add a microcontroller, solenoid, motor speed controller, a button, and code to fire 30 seconds after the button's press. As several members have experience with using microcontrollers (both Circuit Python and Arduino) and we all have significant experience with the hardware after this term we are confident this would not be challenging to implement.

Air Cylinder Improvements: We limited our max air pressure in our tests to 100psi even though our cylinder was rated to 250 due to safety concerns; by using the full pressure that our air cylinder can use, we will greatly increase the distance. We also plan on drilling out the exit shaft of the air cylinder to allow the air to escape faster which would result in faster final velocities without needing an increase in air pressure. Together these two improvements (higher pressure and a more efficient design) will greatly increase the range of the Frisbee Flinger.

High Flow Solenoid: Although adding a remotely controlled solenoid valve was already mentioned under Automation, it has another benefit. A High Flow Solenoid valve would allow us to take advantage of a higher from our air source immediately whereas our current ball valve takes a fraction of a second to ramp up to its maximum flow rate. In normal use of a ball valve amount of time is not significant, but as we seek maximum efficiency and release velocity, every small improvement helps.

Angle Control: By adding an adjustable base similar to a camera bipod or tripod, we will allow the user to set the angle of release and therefore the shape of the throw.

Spin Control: Using the microcontroller and speed controller, we will add a dial to allow the user to control the speed and direction of the disc's spin.

Disc Magazine: We believe that a magazine of several discs is the last improvement we will need to make before releasing this product to market. After a magazine is added, the Frisbee Flinger will reach its full potential, allowing a single ultimate player to practice catching a throw multiple times in a row without a partner.

Business Plans

Though Ultimate is a young sport, businesses centered around high level Ultimate training like Greatest Sky Trainers, CUT Camp, Excel Ultimate, and Hive Ultimate show that there is a market for this type of product. Currently these all of these businesses center around selling a digital workout plan or other Frisbee related resources. The only physical product in this space is the Greatest Sky Trainer which does not simulate throws, but can only hold a stationary frisbee on the end of a pole



Videos

IMG_1806.mov

Launching with the bearings spin mechanism

IMG_1812.mov

Launching with the modified (low-friction) spin mechanism