Project

# Title Team Members TA Documents Sponsor
32 A Wearable Device That Can Detect Mood
 Junjie Ren
Kejun Wu
Peidong Yang
Xinzhuo Li
 design_document1.pdf
design_document2.pdf
final_paper1.pdf
final_paper2.pdf
proposal2.pdf
proposal1.pdf
 Said Mikki
# A Wearable Device That Can Detect Mood

**Team Members:**
- Junjie Ren [junjier2]
- Peidong Yang [peidong5]
- Xinzhuo Li [xinzhuo4]
- Kejun Wu [kejunwu2]

## Problem
Our project targets the pervasive impact of workplace stress, anxiety, and depression, recognizing these as critical challenges compromising individual well-being and overall productivity. Motivated by the need for proactive solutions, we aim to provide a wearable device equipped with advanced sensors and a unique mood recognition framework. By integrating psychological knowledge and wearable technology, our solution objectively monitors and manages mood-related challenges, offering timely feedback. The goal is to contribute to a healthier work environment, and our project represents a significant step at the intersection of technology and mental health in modern workplaces.

## Solution Overview

### Objective
The project aims to recognize and monitor the mood of employees in a workplace environment, leveraging wearable sensors and smartphone technology.

### Problem-Solving Approach
1. **Mood Recognition**: Using wearable sensors to collect physiological data that correlates with various mood states.
2. **Data Analysis**: Applying machine learning algorithms to interpret the physiological data and predict mood states.
3. **Feedback Mechanism**: Providing individual feedback to users and aggregated data to employers for well-being initiatives.

## Solution Components
1. **Wearable Sensor Subsystem**:
- **Components**: Practical sensors, like Toshiba Silmee™ Bar Type or W20/W21 wristbands.
- **Function**: Collects physiological data such as heart rate, skin temperature, and activity levels.
- **Role in Solution**: Provides the raw data necessary for mood prediction.

2. **Data Processing and Analysis Subsystem**:
- **Components**: Machine learning models (both personalized and generalized), feature extraction techniques.
- **Function**: Analyzes sensor data, extracts meaningful features, and applies machine learning techniques to predict mood.
- **Role in Solution**: Core of the mood recognition framework, turning data into actionable insights.

3. **Feedback and Reporting Subsystem**:
- **Components**: User interface for feedback, anonymized data aggregation for employers.
- **Function**: Provides mood predictions and wellness statistics to users and employers.
- **Role in Solution**: Closes the loop by informing users about their mood trends and assisting employers in enhancing workplace wellbeing.

## Criterion for Success

### Hardware Achievements
1. Successful deployment of advanced sensors, which are capable of collecting various physiological data such as heart rate, pulse rate, and skin temperature.
2. Integration of sensors into a wearable format that can be comfortably used in the working environment.
3. Clear and vivid display on the screen, indicating the detected mood.

### Software Achievements
1. Creation of a sophisticated mood recognition framework capable of identifying eight different types of moods at five intensity levels, with regular time updates.
2. Application of machine learning techniques for both personalized and generalized mood prediction models based on physiological data.
3. Achievement of a high average classification accuracy in mood prediction, showcasing the efficacy of the software algorithms.

## Distribution of Work
- Junjie Ren is responsible for System Design and Architecture.
- Peidong Yang is responsible for Data Collection and Analysis.
- Xinzhuo Li takes charge of Psychological Model Integration.
- Kejun Wu is responsible for User Study Coordination.

### Electrical Complexity
The project's electrical complexity encompasses integrating advanced sensors into the wearable device, demanding intricate signal processing algorithms, and robust coding for accurate mood interpretation. Ensuring seamless communication with the smartphone app adds complexity, along with implementing an efficient power management system for sustained monitoring.

### Mechanical Complexity
The mechanical intricacy involves designing a comfortable and durable wearable, accommodating integrated sensors while considering user ergonomics. The challenge lies in achieving a balance between functionality and aesthetics, ensuring the device is robust enough for daily wear and capable of withstanding various environmental factors. This complexity is justified by the need for a reliable, user-friendly solution contributing to mental health monitoring in professional settings.

Assistive Chessboard

Featured Project

Problem: It can be difficult for a new player to learn chess, especially if they have no one to play with. They would have to resort to online guides which can be distracting when playing with a real board. If they have no one to play with, they would again have to resort to online games which just don't have the same feel as real boards.

Proposal: We plan to create an assistive chess board. The board will have the following features:

-The board will be able to suggest a move by lighting up the square of the move-to space and square under the piece to move.

-The board will light up valid moves when a piece is picked up and flash the placed square if it is invalid.

-We will include a chess clock for timed play with stop buttons for players to signal the end of their turn.

-The player(s) will be able to select different standard time set-ups and preferences for the help displayed by the board.

Implementation Details: The board lights will be an RGB LED under each square of the board. Each chess piece will have a magnetic base which can be detected by a magnetic field sensor under each square. Each piece will have a different strength magnet inside it to ID which piece is what (ie. 6 different magnet sizes for the 6 different types of pieces). Black and white pieces will be distinguished by the polarity of the magnets. The strength and polarity will be read by the same magnetic field sensor under each square. The lights will have different colors for the different piece that it is representing as well as for different signals (ie. An invalid move will flash red).

The chess clock will consist of a 7-segment display in the form of (h:mm:ss) and there will be 2 stop buttons, one for each side, to signal when a player’s turn is over. A third button will be featured near the clock to act as a reset button. The combination of the two stop switches and reset button will be used to select the time mode for the clock. Each side of the board will also have a two toggle-able buttons or switches to control whether move help or suggested moves should be enabled on that side of the board. The state of the decision will be shown by a lit or unlit LED light near the relevant switch.