Muvik | Devpost

Inspiration

The community of Boston is a vibrant hub for runners, yet a recurring question persists: "What music are people listening to?" Traditional music suggestions often feel stagnant; a Spotify playlist with 100,000 saves frequently contains tracks that users skip, ultimately disrupting the momentum of a run. We developed Muvik to ensure runners never have to manually adjust their music, providing a seamless transition between their physical pace and the emotional "vibe" of their exercise.

What It Does

Muvik is a wearable device that utilizes an accelerometer to track a user’s walking or running cadence in real-time. By analyzing movement frequency, the system curates a musical selection at the appropriate Beats Per Minute (BPM). As the runner accelerates or decelerates, the music dynamically adjusts to match their current tempo.

How We Built It

The project architecture integrates hardware-level processing with a modern web interface:

Front-End: Developed using HTML and CSS to create a dashboard that displays the real-time BPM of the active track.

Back-End: Written in C for high-efficiency sensor data processing and management of the LCD display output.

Hardware: Centered around an ESP32 microcontroller, an Accelerometer for gait tracking, and an LCD Display to visualize tempo and mood metrics.

The Logic Behind the Music

To convert raw accelerometer data into a musical tempo, we implemented a calculation to determine the runner's cadence. We capture the number of steps detected by the sensor over a specific time interval and multiply it to find the target BPM. To prevent the music from "jittering" during sudden arm movements, we applied a smoothing algorithm that averages new data with previous readings to keep the transitions fluid.

Challenges We Ran Into

Our initial design phase focused on utilizing blood oxygen and heartbeat sensors as primary biometrics. However, we discovered that these sensors lacked the necessary accuracy for high-intensity movement or failed to provide consistent data in a mobile environment. This led us to pivot toward accelerometer-based gait analysis, which proved much more resilient.

Accomplishments That We're Proud Of

We are particularly proud of our ability to maintain our core vision despite significant hardware setbacks. Successfully pivoting from biometric sensors to motion-based tracking allowed us to deliver a functional prototype within the hackathon's time constraints.

What We Learned

This project reinforced a fundamental engineering mantra: "If there is a sensor, there is a way." We learned to value the reliability of motion data over the complexity of biological data in wearable contexts.