Jupiter Squat System
Monitoring Squat Depth. Improving Performance.
Overview
Key Challenges
Weekly Design Reviews with Investor
Mapping Infrared Light
Attaching Sensor to Clothing
My Role
Mechanical Engineer - Team of 2
Talo Fitness LLC
Boston, MA
Skills Applied
CAD Modeling (SolidWorks)
CNC Mill
Laser Cutting
3D Printing
Arduino
Soldering
Customer Segmentation
User Experience and Testing
Year
May 2019
About
The Jupiter Squat System provides coaches and athletes an objective and consistent way to monitor squat depth. Coaches are able to monitor multiple athletes at once to ensure training leads to improved performance.
Consistently performing squats correctly leads to improved performance in competition and develops more muscle to help improve overall strength and health.
System Overview
From Proof of Concept to Alpha
Proof of Concept
-
60% Accuracy
-
1 Unit
-
Quick assembly of device and key components
-
Demonstrated feasibility
-
Presentation to Investor
Alpha
-
80-100% Accuracy
-
2 Units
-
Look-like Works-like product
-
Tested and evaluated design to repair flaws
-
Conducted limited user testing
The first step in developing Alpha was figuring out a way to package the components in a way that would create a more refined look and improved stability in the stand. I began by sketching the layout of components and quickly realized that they could be stacked into two compartments.
The Alpha stand was fabricated in-house. I used a 3D printer, CNC mill, and laser cutter to create all of the custom parts. The remaining components were ordered OTS.
First Round Testing
Summary
Testing of the Alpha prototype included user experience testing and product accuracy testing. In total, 4 users completed 10 rounds of testing each.
Users:
User 1: Male, 6’0”, 22, Expert squatter, Back squat (high bar)
User 2: Male, 5’8”, 22, Proficient squatter, Back squat (in between)
User 3: Male, 6’6”, 22, Proficient squatter, Back squat (unknown)
User 4: Female, 6’0, 25, Novice squatter, Back squat (unknown)
Results
-
60% Accuracy in detecting target squat depth.
-
100% of users found it "extremely valuable" to know when they've hit their target depth.
-
95% Comfort: users felt that the hip unit was fairly comfortable.
-
95% Intuitive Set up: users found it easy to adjust stand to proper height.
-
93% Audio Feedback: users found the audible alert valuable.
Troubleshooting
Before proceeding with further testing, we needed to improve the accuracy of the device. The first round of testing had many false positives and negatives. It appeared that the IR signal was scattering and being picked up by sensor at any point during the squat. We found a way to map the IR light by sensing and marking the upper and lower bounds of the signal.
The IR beam that was being used on the device was an inexpensive laser line. The mapping test we ran quickly showed us that the IR signal was scattering up to 10 inches away from the line. This explained all of the false positives. We decided to purchase a more expensive IR laser line and repeated the test. The results drastically improved.
Second Round Testing
Summary
Testing of the Alpha prototype included user experience testing and product accuracy testing. A more accurate laser line was installed for the test. In total, 6 users completed 10 round of testing each.
Users:
User 1: Male, 5’6”, 79, Proficient squatter, Back squat (in between)
User 2: Male, 5’8”, 79, Proficient squatter, Back squat (in between)
User 3: Female, 64, 5’11”, Proficient squatter, Back squat (in between)
User 4: Male, 5’11”, 50, Expert squatter, Back squat (Low bar)
User 5: Female, 4’11”, 49, Proficient squatter, Back squat (in between)
User 6: Female, 5’2”, 70, Proficient squatter, Front squat
Results
-
80-100% Accuracy in detecting target squat depth.
-
100% of users found it "extremely valuable" to know when they've hit their target depth.
-
93% Comfort: users felt that the hip unit was fairly comfortable.
-
83% Intuitive Set up: users found it easy to adjust stand to proper height.
-
93% Audio Feedback: users found the audible alert valuable.
Iterating on the Hip Unit
The results of the second round of testing improved from the first. However there continued to be some instances of false positives and negatives. This time we decided to look at issues with the hip unit instead of the stand. As we replayed the testing videos, we realized that as the user squatted, the hip unit would rotate, and the IR sensor would no longer be measuring at the hip crease.
We decided to reduce the weight of the hip unit by reducing the size of the components. I designed an acrylic enclosure for an Arduino nano and a smaller breadboard.
The second version of the hip unit only had to be tested once to find out that while the smaller size helped, the sensor was still rotating upon squatting. We began to brainstorm new ways to attach the IR sensor to a user's hip crease.
