1 of 1

RiseVBT

F4: Jason Botros, Jamshed Panthaki, Meadow Webster

18-500 Capstone Design, Spring 2025

Electrical and Computer Engineering Department

Carnegie Mellon University

Product Pitch

Accurate assessment of athletic performance is critical for optimizing strength training and reducing risk of injury, yet most current methods often rely on subjective observations or limited quantitative measurements. Coaches and athletes need a reliable, easy-to-use system that provides real-time metrics to monitor training sessions and make data-driven programming decisions.

RiseVBT presents a novel velocity- based training system that integrates an IMU sensor with an iPhone camera to quantify athletic performance by measuring bar velocity and orientation. By fusing sensor data with a visualization of the traveled bar-path, the proposed system provides a more robust analytical framework than conventional methods involving standalone sensors, thereby contributing to the enhancement of performance monitoring and training efficacy in strength sports.

The system is required to automatically store distinct data on each rep in a set for key barbell lifts (squat, deadlift, bench, overhead press), provide real-time objective feedback with velocity and orientation visualizations, allow for easy setup with lightweight and discrete gear that is compatible with any barbell, and function accurately in standard gym conditions.

Our mobile app is the origin of RiseVBT’s operation, where users store data. A camera API handles detection for bar path tracking and Bluetooth API links the iOS app to the hardware system. Velocity and orientation data is collected from the sensor during lifts, which is seamlessly transmitted back to the app through an ESP32 for a post-lift report to the user. Battery levels and calibration status are reported on a TFT display for easy monitoring of device status.

Our integrated device is able to successfully record sets and provide corresponding feedback that displays accurate scientific analysis. However, minor cases exist that show drift in velocity data due to falling out of full calibration and the orientation of the sensor dictates how output data is interpreted. But overall, RiseVBT accomplishes our goal to provide objective feedback to weightlifting through a scientific lens. Future implementations have the opportunity to apply customization tailored to a user’s performance that suggests lifts based on data history, elevating the visual feedback system to give users a way to critique their form relative to a model movement, or adjusting the sensor to function equally with any orientation.

RiseVBT operates as a joint hardware-software system. The iOS application handles data storage, camera API, Bluetooth API, and the overall UI. The physical sensor assembly includes a BNO055 IMU, 3.3V Li-Po battery, and an ESP32-S3 TFT Feather, which controls Bluetooth to link the two systems.

Hardware Design

Velocity vs. Time in 5 Rep Benchpress

Independent hardware and software unit tests were conducted iteratively throughout the semester, with separate full-system tests following integration. The following tests validate the device’s performance within the previously defined use-case requirements.

Technical Requirements:

Software Workflow

BNO055 Sensor

ESP32

Battery

Barbell Clip

TFT Display

System Description

System Evaluation

System Architecture

Conclusions & Additional Information