Hardware and Software Layers

Buddi’s design relies on a modular, multi-layered architecture that separates hardware sensing components from software intelligence, ensuring scalability, adaptability, and privacy-conscious design. This separation also allows for easy upgrades or extension with new devices or algorithms without affecting the overall ecosystem.

1. Hardware Layers

The hardware layer is designed to capture emotional, behavioral, and engagement data in a non-intrusive, classroom-friendly manner.

A. Ambient Sensors

  • Purpose: Detect subtle cues from students such as motion, posture, or environmental context.
  • Examples:
    • Desk-mounted light & proximity sensors
    • Ambient microphones for noise-level monitoring (privacy-preserving)
    • Optional temperature, light, and CO₂ sensors to monitor classroom comfort
  • Role: Feed real-time context to the Companion Intelligence Layer to enhance emotional inference.

B. Wearable Devices (Optional)

  • Purpose: Capture physiological signals linked to emotional states.
  • Examples:
    • Heart rate or pulse sensors
    • Skin conductance / galvanic sensors
    • Lightweight, wrist-worn or clip-on devices
  • Role: Provide fine-grained, real-time affective data for companion adaptation.

C. Student Interaction Devices

  • Purpose: Facilitate direct engagement with Buddi’s virtual companion and gamified modules.
  • Examples:
    • Tablets, laptops, or interactive touchscreens
    • Keyboards and mouse input for engagement tracking
  • Role: Deliver micro-interventions, feedback, and gamified experiences.

2. Software Layers

The software layer is the intelligent core of Buddi, responsible for emotion recognition, adaptive feedback, gamification, and classroom analytics.

A. Companion Intelligence Layer

  • Components:
    • Emotion Recognition Engine: Processes sensor input to infer emotional states.
    • Behavior Modeling Module: Maintains dynamic profiles for adaptive responses.
    • Adaptive Response Generator: Determines the companion’s expressions, micro-interventions, and encouragement prompts.
  • Tech Stack: Python / Node.js, TensorFlow Lite / PyTorch Mobile, REST/WebSocket APIs.

B. Intervention & Engagement Layer

  • Components:
    • Calm Kit: Sensory-friendly, ultra-short interventions (breathing exercises, doodling, grounding gestures).
    • Gamification Engine: Soft progress visualizations, constellations, and reward loops.
    • Focus Trails: Tracks engagement and transforms study patterns into interactive visual feedback.
  • Role: Sustain engagement, promote self-regulation, and provide playful reinforcement.

C. Data Aggregation & Analytics Layer

  • Components:
    • EchoGarden: Aggregates classroom mood data in real-time, anonymized for privacy.
    • Analytics Engine: Generates trends, reports, and insights for educators.
    • Secure Storage: Encrypts all data, ensuring privacy and compliance.
  • Tech Stack: PostgreSQL / Firebase, Python analytics libraries, D3.js / Plotly for visualizations.

D. User Interaction Layer

  • Components: Frontend interface, mobile/tablet apps, dashboards.
  • Tech Stack: React Native / Flutter, Material Design, CSS/Sass for styling.
  • Role: Ensure intuitive, accessible, and low-stimulation experiences for neurodivergent learners.

3. Integration Between Hardware and Software

  • Data Flow: Sensor & interaction data → Companion Intelligence Layer → Intervention & Engagement → Analytics Layer → Feedback to user & educator dashboards.
  • Local Processing: Emotion recognition can occur on-device, minimizing latency and preserving privacy.
  • Cloud/Server Support: Aggregated, anonymized data stored for classroom insights, historical trends, and adaptive learning improvements.

4. Key Design Principles

  1. Non-Intrusive: Sensors and wearables are optional and minimally invasive.
  2. Scalable: Modular hardware and software layers allow easy addition of new sensors, interventions, or classrooms.
  3. Privacy-Focused: Personal data never leaves the device unless anonymized for classroom analytics.
  4. Adaptive: Companion behavior, interventions, and gamification adapt over time based on user engagement and emotional feedback.

In summary, the hardware and software layers of Buddi work in harmony to create an emotionally intelligent, interactive, and privacy-conscious ecosystem that supports neurodivergent learners and empowers educators with actionable, anonymized insights, all in a playful and accessible environment.

Next : Workflows