Shingle Shark

Figure 1: Top-View Into Shingle Shark Main Body

Project Summary: Design of a low-weight, low-cost, electromechanical roof-shingle removal device capable of operating on high pitched roofs.

Team

Jacky Hua

Freddy Ramon

Jesse M. Okoche

Nico Ramos

Abstract

Asphalt shingle removal is labor-intensive, especially on steep roofs. Current shingle-removal appliances are typically heavy, requiring pneumatic pistons and a separate compressor to create the force required to rip up shingles and nails. Our capstone project, the Shingle Shark, aims to create a lighter, more compact electromechanical shingle-removal device that reduces user effort and improves usability on steeper roof pitches. The device uses a motor-driven four-bar linkage to convert rotation into a scooping-and-lifting motion. This motion then drives a low-carbon steel head under shingles to lift them out. The final integrated prototype uses a pre-geared 90 V DC motor, bevel-gear power transmission, aluminum linkage components, a steel head, rear wheels, and a one-handed handle. Through three prototype iterations, we validated the linkage motion, demonstrated nail removal using a full-scale hand-cranked proof of concept, and produced an electrically powered prototype capable of single- and double-nail removal. Testing also revealed limitations in our design, including gear-meshing inconsistency, stand deflection, gear wear, and crank-arm wear. The prototype demonstrates a feasible low-cost alternative to pneumatic shingle-removal tools. Future work should focus on improving drivetrain robustness and longevity, gear upgrades, tolerance control, fatigue testing, and rigorous field testing.

Introduction: State of the Art Benchmarking

Commercial shingle-removal products such as the Rapid Roof Remover and Shingle Hog provide mechanized assistance for roof tear-off work. These tools demonstrate that powered shingle removal is feasible, while also revealing several limitations that motivated this project. Existing products are often pneumatic, require a compressor, and are heavy or bulky [1] [2]. Because of their weight and being able to stand for operation, they are only practical on walkable roof slopes (up to 6/12 pitch – or 22° slope).

Figure 2: Rapid Roof Remover

Figure 3: Shingle Hog

Benchmarking these products led to the following design targets for Shingle Shark, reflecting the functional need to remove shingles and the human factors needed to reduce roof-worker fatigue and instability:

Table 1. Shingle Shark Design Targets Based on Rapid Roof Remover and Shingle Hog Standards

Specification	Target
Weight (lb.)	< 30
Operation	Electromechanical
Tooth-bar/Blade Width (in.)	12-18
Roof Pitch Range	< 8:12
Shingle Removal Rate (square/hour)	20
Ergonomics	One-handed
Dimensions (L x W x H in.)	42 x 20 x 30

Methodology: Design Process and Prototype Development

We followed an iterative build-test-learn process. Rather than beginning with a fully integrated machine, the design was divided into functional modules and matured through three prototype stages:

Prototype 1: Small-scale motion validation.

Figure 4: MotionGen Four-Bar-Mechanism Generated from Desired Coupler Path Motion

Figure 5: Prototype 1 - 3D Printed PLA Four-Bar Linkage from MotionGen in Isometric View

Prototype 1 successfully demonstrated the desired motion. The linkage moved smoothly, did not lock up, and generated a motion that passes close to the roof’s surface. This validated the use of a four-bar linkage for the project and justified moving to a full-scale mechanical proof of concept.

Prototype 2: Full-scale metal proof of concept.

Figure 6: Hand-crank-driven, Machined, Steel Proof-of-concept Prototype 2

Prototype 2 successfully removed a roofing nail, proving that the four-bar concept could perform useful work at full scale. However, testing also identified several issues:

The scooping angle was not optimal.
The full-scale prototype was heavy and bulky.
The teeth experienced excessive bending.

These findings informed the next design iteration. We altered shaft stand heights to change the motion, removed unnecessary material in our linkages to reduce weight and modified the tooth geometry and material.

Prototype 3: Fully integrated electrically powered prototype.

Internal Components of Shingle Shark Device

Figure 7: Powertrain-integrated, Full-scale

Final Prototype Architecture

The final Shingle Shark structure consists of the following major subsystems, partly observed in Figure 7 above:

Parallel four-bar linkage motion-generation system.
Low-carbon-steel tooth-bar.
Pre-geared DC motor.
Bevel-gear power transmission.
Central driveshaft
Optimized aluminum crank and rocker arms and base plate.
Gearbox and protective housing.
Rear-wheel mobility and traction system.
Trimmer-style handle and user controls.
A guarded housing with a clear acrylic viewing window.

The motor rotates a power shaft through a set of bevel gears. The rotating shaft drives two crank arms, connected to a set of parallel four-bar linkages. The linkage generates a scooping motion that moves the toothed head forward beneath shingles, and then upward, prying nails up.

Prototype 3 successfully removed nails using motor power and demonstrated the feasibility of the integrated Shingle Shark architecture.

Table 2. Prototyping Summary Table with Purpose, Key Results, and Limitations

Prototype	Purpose	Key Result	Main Limitation
1	Small-scale motion validation	Smooth scoop-and-lift motion validated	Not load-bearing
2	Full-scale metal proof of concept	Successful hand-cranked nail removal	Heavy, bulky, tooth-bending
3	Fully integrated powered device	Successful motor-powered nail removal	Gear meshing, shaft drift, wear, stand deflection

Test Plans and Instrumentation

Nail-Removal Force Requirement

We performed nail-removal force testing to establish drivetrain and structural requirements. A roofing nail was removed using a prybar which was attached separately to a spring scale. The experimental force required to remove a single nail was approximately 60 lbf. The final presentation identified the target lifting capacity as approximately 124 lbf, incorporating a 25% contingency factor and a 1.47 rust factor. This value was used as a conservative basis for mechanism and drivetrain design [9].

Figure 8: Test-rig to find the force to remove a nail

The initial nail-pull test was used in FEA models to create and size the components for our final device. A final experiment done to test the actual force output of the device is required to verify the tests, models, and assumptions used.

Strain Gauge Testing

A 350 Ohm strain gauge was used as one of the 4 resistors in a Wheatstone Bridge, placed at the center of the head where we expect the most bending to take place. A LabJack T7 is used to measure the voltage across the center of the Wheatstone Bridge and outputs along with a counter that measures time to an output file. 350 Ohm resistors weren’t accessible, so a 360 Ohm and 13000 Ohm resistor were placed in parallel so that an equivalent 350 Ohm resistor would be placed at each of the nodes. That way, the effect of the DC Offset in the LabJack’s final measurement would be greatly minimized. To keep the wires orderly, black wires are for ground, red wires are hot, white wires represent the middle wires from the right half of the diagram, and green wires represent the middle wires in the left half of the diagram.

Figure 9: Circuit diagram and real application of the strain gauge wiring.

Calibration was performed by applying known weights to the head and measuring the corresponding voltage. A total of 4 weights were placed on the head. The change in voltage, while measurable, took place in the order of millivolts, so a pre-built amplification function in the LabJack scaled the measurement to tenths of volts for data processing. Fluctuations are natural on the order of millivolts, so all calibration measurements were taken over the span of a minute so that an average of the voltage could be obtained for a given weight. The calibration data is shown below:

Figure 10: Strain gauge calibration data for 3 data points.

A linear fit was used to map the force output to the voltage. While a single data point is not enough to characterize deflection, it’s clear that a rolling average shows an effect from the additional weight to the voltage. With this calibration, a single-nail removal test was done to validate the force to pull a nail. The results are shown below:

Figure 11: Voltage output for a single nail extraction

During a single-nail removal test, the strain-gauge measurement was approximately 0.32 V, which corresponds to a 6.1 lbf nail removal force. The reason for such a low force was for safety – the strain gauge and the LabJack testing setup had lots of loose wires that connected laptops, LabJacks and the rig together, and the strain gauge was fully exposed on the head of the Shingle Shark. Therefore, in our test, we used wood half as thick as what roofers might nail shingle in, and we only nailed the nail halfway to try to take data. Even so, the test confirmed that the instrumentation could capture meaningful load-related data during operation. The final testing setup is shown below:

Figure 12: Strain-gauge and LabJack T7 testing setup

Future Work

Upgrade the gear set: Replace the current miter bevel gears with a stronger gear set, such as helical bevel gears. This would improve contact and bending safety factors and reduce wear.
Improve gear alignment and tolerance control: Add more rigid bearing supports, improve axial shaft constraints, tighten keyway tolerances, and perform a formal tolerance stack-up analysis for the drivetrain.
Reinforce the rocker stands: Add reinforced brackets, vertical bolting, or a stiffer stand geometry to reduce deflection under load.
Improve crank-arm durability: Explore changing crank-arm material from aluminum to steel or redesigning the crank arms to reduce wear in the key-hole.
Refine the head wedge and tooth geometry: Continue testing different tooth angles, taper profiles, and head widths to improve nail engagement while minimizing tooth bending and roof-deck damage.
Conduct systematic durability testing: Test repeated cycles under controlled conditions to measure wear, drift, jam frequency, and failure modes.
Quantify performance metrics: Future testing should document extraction success rate, cycle time, removal rate, current draw, substrate damage, and user effort.
Test on realistic roof assemblies: Begin with controlled mock-roof testing at multiple slopes, then consider real-roof testing only after safety and reliability are improved.
Improve production-intent safety features: Replace prototype viewing features with durable guarding, add fail-safe shutoff behavior, and develop a formal safe operating procedure.
Investigate additional weight reduction: Reduce the motor stand size, optimize the housing, and consider alternative lightweight materials only where stiffness and durability are preserved.

References

“Home | The Shingle Hog,” The Shingle Hog. Accessed: May 6, 2026. [Online]. Available: https://www.shinglehog.net/
“Rapid Roof Remover | Pneumatic shingle and flooring removal tool,” Rapid Roof Remover. Accessed: May 6, 2026. [Online]. Available: https://www.rapidroofremover.com/
T. M. Johnson, “Power tool system and method for removing roof shingles,” U.S. Patent Application Publication US 2023/0003030 A1, Jan. 5, 2023. [Online]. Available: https://patentimages.storage.googleapis.com/7b/08/b8/c2695048e6fbf6/US20230003030A1.pdf
D. C. Willis, “Rapid roof remover,” U.S. Patent US 7,313,985 B2, Jan. 1, 2008. [Online]. Available: https://patents.google.com/patent/US7313985B2/en
J. K. Nisbett and R. G. Budynas, Shigley’s Mechanical Engineering Design, 2024 release. New York, NY, USA: McGraw Hill LLC, 2024. ISBN: 978-1-265-47269-6.
D. G. Ullman, The Mechanical Design Process, 4th ed. New York, NY, USA: McGraw-Hill Education, 2009. ISBN: 978-0-07-297574-1.
American Gear Manufacturers Association, Fundamental Rating Factors and Calculation Methods for Involute Spur and Helical Gear Teeth, ANSI/AGMA 2001-D04. Alexandria, VA, USA: American Gear Manufacturers Association, 2004.
McMaster-Carr, “Motor, bevel gear, shaft, bearing, and hardware specifications used for component selection,” McMaster-Carr. Accessed: May 6, 2026. [Online]. Available: https://www.mcmaster.com/
R. Takanashi, K. Sawata, Y. Sasaki, and A. Koizumi, “Withdrawal strength of nailed joints with decay degradation of wood and nail corrosion,” Journal of Wood Science, vol. 63, pp. 192–198, 2017, doi: 10.1007/s10086-016-1600-5.

Figure 13. 2025-26 Mechanical Engineering Capstone Poster Session Contribution