Tech-Neck Feedback to Improve Homework Posture

ISEF Category: Translational Medical Science

Subcategory: Disease Prevention  ·  Difficulty: Intermediate  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

A lot of students spend hours staring down at a screen. That head-down posture can stack up like interest on a loan, one small slump at a time. Your project turns that hidden load into a number you can track. Then you can test whether a phone vibration helps people sit up straighter.

What Is It?

This project measures how much time your neck spends bent forward during homework, test prep, or screen time. Think of it like an activity score for posture. Instead of counting steps, you count neck flexion, which means how far your head tilts forward from a neutral position.

A webcam and a pose-tracking model, such as MediaPipe, can estimate body landmarks from video. From those landmarks, you can calculate a simple posture score over time. If the score stays high for long stretches, that suggests more cumulative strain, or what you can call tech-neck dose. Your question is whether a quick vibration from a phone can interrupt that pattern and lower the total dose.

This is a prevention project, not a diagnosis project. You are not trying to prove injury. You are testing whether a feedback cue changes behavior in a measurable way.

Why This Is a Good Topic

This is a strong science fair topic because you can measure a real behavior, change one thing, and compare before-and-after data. It connects to a common problem, screen-related neck strain, and it gives you a clear intervention, haptic feedback. You can learn pose tracking, data logging, signal processing, and basic statistics without needing a hospital or a lab.

Research Questions

• How does real-time phone vibration affect cumulative neck-flexion dose during homework sessions?
• What is the effect of haptic feedback on the number of long forward-head posture episodes per session?
• Does a webcam-based posture score agree with a self-reported discomfort score after study sessions?
• To what extent does feedback reduce tech-neck dose in the first week compared with the fourth week?
• Which study setup, desk, couch, or bed, produces the highest cumulative neck-flexion dose?
• What is the effect of alert threshold choice on false alarms and posture improvement?
• How does session length change the total neck-flexion dose with and without feedback?

Basic Materials

• Laptop or desktop computer with a working webcam.
• Smartphone with vibration alerts and a timer or notification app.
• Chair, desk, and a consistent study setup.
• Free MediaPipe-compatible coding environment, such as Python on your computer.
• Tripod, stack of books, or laptop stand to keep the webcam stable.
• Notebook or spreadsheet for session logs and self-reported discomfort ratings.
• Consent form for any family member or classmate who helps test the system.

Advanced Materials

• External webcam with fixed frame rate.
• Optional depth camera or second camera for comparison.
• Surface EMG sensor for exploratory neck muscle activity if a lab has one.
• Goniometer or inertial sensor for validation against webcam angle estimates.
• Data-logging phone app or Bluetooth trigger for feedback timing.
• Python environment with pose-estimation, data-cleaning, and plotting libraries.
• Statistical software for repeated-measures analysis.

Software & Tools

• Python: Runs the pose-tracking code, the dose calculation, and the data analysis.
• MediaPipe: Detects body landmarks from webcam video so you can estimate neck angle.
• ImageJ: Helps you inspect frames and compare posture landmarks if you need visual checks.
• Google Sheets: Tracks session dates, feedback condition, and self-reported discomfort scores.
• JASP: Runs basic statistics, including paired tests and repeated-measures comparisons.

Experiment Steps

1. Define your posture signal, such as neck angle, forward-head threshold, or cumulative time above a cutoff.
2. Choose one feedback rule and keep it fixed, such as vibration after a posture threshold is crossed for a set pattern.
3. Plan a control condition that matches study time, but gives no feedback.
4. Build a simple calibration method so your webcam score maps to the same body position across sessions.
5. Decide how you will log dose, discomfort, and session context for every trial.
6. Pre-plan your statistics so you compare each person to themselves across feedback and control weeks.

Common Pitfalls

• Letting the webcam move between sessions, which changes the landmark estimates and makes your posture score drift.
• Using a feedback threshold that is so sensitive it vibrates for normal micro-movements, which trains the user to ignore it.
• Mixing study tasks with different head positions, which makes homework type look like a feedback effect.
• Forgetting to validate the pose model against a simple posture check, which can hide angle errors.
• Measuring only one session per condition, which gives you too little data to tell real change from noise.

What Makes This Competitive

A stronger project will do more than count posture changes. You can compare several feedback rules, test whether the effect lasts after the novelty wears off, and report effect sizes instead of only p-values. You can also validate the webcam score against another measurement, which makes your method much more credible. A clear design, careful controls, and honest error analysis can turn a simple idea into a serious research project.

Project Variations

• Test the same feedback system during gaming sessions instead of homework to see whether attention level changes posture response.
• Compare vibration feedback with visual pop-up alerts to see which cue reduces neck-flexion dose more.
• Use a phone accelerometer or wearable sensor instead of webcam pose tracking and compare the two measurement methods.

Learn More

• PubMed: Search review articles on forward head posture, neck pain, and screen-related musculoskeletal risk.
• NIH National Library of Medicine: Look for research on ergonomics, posture, and adolescent musculoskeletal health.
• MediaPipe documentation: Read the official pose estimation guide to learn how landmark tracking works.
• Google Open Source Blog: Find plain-language explanations and examples of MediaPipe computer vision projects.
• MIT OpenCourseWare, Introduction to Computer Science and Data Analysis: Use free course materials for coding, data handling, and visualization.