Emotions (in) Transit

A dialogue between data, body, and city. Master's Thesis (TFM) - Generative Art Installation & Biofeedback.

---

The Urban Paradox: Context

We live in the era of the Smart City: cities optimized by data and efficiency. However, the WHO warns of a growing urban mental health crisis. We are digitally connected, yet emotionally isolated.

This project intervenes in "Non-Places" (Marc Augé's concept): anonymous transit spaces where we coexist but do not connect.

The Central Question

Can design, acting through technology and generative art, create a bridge of empathy within these anonymous spaces?

---

The Ethical Pivot: From Surveillance to Connection

During development, the project evolved drastically based on ethical findings:

Phase 1: Exploration (Discarded)	Phase 2: Final Proposal (Biofeedback)
Use of Computer Vision for facial emotion detection.	Use of Biometric Sensors (pulse) + Environmental Data.
Perceived as Surveillance.	Perceived as Connection.
Focus on "EGO" (identification).	Focus on "ECO" (systemic connection).

---

Technical Architecture

The system synchronizes the "slow" pace of the city with the "fast" pace of the human body through a hybrid architecture:

1. Hardware (Low Latency IoT)

• Brain: ESP32 Microcontroller (Dual Core).
  • Core 0: Wi-Fi and WebSockets management.
  • Core 1: Uninterrupted sensor reading.
• Sensor: MAX30102 (Pulse Oximetry).
• Latency: < 100ms (perceptual real-time).

2. Urban Variables (Data Inputs)

External APIs are integrated to define the city's "mood":

• Acoustic Stress: Noise level (dB) [SmartCitizen].
• Collective Flow: Transport demand [TMB API].
• Atmosphere: Weather and pressure [OpenWeatherMap].

3. The Synchronization Challenge (Python Backend)

How to merge human pulse (10ms) with atmospheric pressure (1h)? We use Pandas and pd.merge_asof to align time series by proximity, allowing the "slow" state of the city to modulate the user's "fast" visual response.

4. Generative Visualization (Frontend)

• Engine: p5.js + WebGL (Shaders).
• System: +4,000 fluid particles at 60 FPS.
• Poetic Logic:
  • High Noise + Low Pressure → Agitated particles ("Urban Anxiety").
  • Fluid Traffic + Stable Weather → Harmonic particles ("Calm").
  • User Pulse → Acts as an anchor and modulator of the chaos.

---

Installation & Deployment

Prerequisites

• Node.js (v16+)
• Python (3.8+)
• ESP32 Device (for the full experience)

Local Setup

1. Clone the repository:

   git clone [https://github.com/Raquel-bena/Emotions-in-Transit.git](https://github.com/Raquel-bena/Emotions-in-Transit.git)

2. Environment Configuration: Rename .env.example to .env and add your API Keys (TMB, OpenWeather).
3. Run Frontend:

   npm install
   npm run dev

4. Run Python Server:

   cd server
   pip install -r requirements.txt
   python app.py

---

Future & Scalability

• Node Network: Installation of multiple totems to compare the "emotional state" across different neighborhoods.
• New Variables: Air Quality (PM2.5) and Social Media Sentiment Analysis.
• 5G Integration: Greater autonomy for installations in public spaces.

---

Links & Resources

• Live Demo: Emotions in Transit (Render)
• p5.js Collection: Editor Sketches
• Portfolio: Instagram @rb.graphicx

---

Developed by Raquel Benavides