PulseBoard is a full-stack, modular, and cloud-native IoT data monitoring platform designed to ingest, store, analyze, and visualize real-time sensor data for multiple organizations. It includes support for AI/ML-based anomaly detection, user and organization mapping, and live dashboards built using modern technologies.

• Sensor Simulation: Generates simulated sensor data (temperature, humidity, pressure, etc.) with realistic edge cases and noise.
• Kafka + Kafka Topics: Real-time data streaming per organization and sensor type.
• Kafka Streams: Filtering, transformation, enrichment of messages.
• TimescaleDB: Time-series storage of sensor telemetry data.
• Redis: Caching of latest sensor readings for quick access.
• GraphQL API: A single flexible endpoint to fetch live and historical data.
• React + Tailwind Frontend: Real-time charts and visualizations.
• Authentication & Organization Management: User registration with org selection and access control.

• Forecasting with ARIMA / Prophet models.
• Threshold-based anomaly scoring based on user-defined limits.
• Adaptable to different sensor types and edge cases.

• Prometheus: Metrics collection from services.
• Grafana: Visual dashboard for backend service health and metrics.
• Custom dashboards for users to monitor sensor data.

Main Components:

• Sensor Simulator → Kafka Broker → Kafka Topics → Kafka Streams → TimescaleDB
• Kafka Streams also forwards to Redis (latest cache) and AI/ML Agent (forecasting)
• Prometheus scrapes TimescaleDB for Grafana metrics
• GraphQL API fetches from Redis + TimescaleDB → React Dashboard (with filtering, live view, history)
• Auth & Org Service handles user-org relationships

• Missing values (nulls)
• Sensor spikes/dips
• Overlapping timestamps
• Sensor-specific anomalies (e.g., negative pressure)
• Thresholds set by users per sensor

• Users register with their email and select from predefined organizations.
• Sensors are mapped to organizations and can be subscribed to by users.
• Each user has access to only their org’s Kafka topics and dashboards.
• Sensor settings are user-specific: enable/disable, thresholds, preferences.

1. Register/Login with selected organization.
2. Sensor Dashboard: Enable/disable sensors, set thresholds.
3. Live Data: Realtime stream charts for selected sensors.
4. Historical Data: Time-based filtering and visualization.
5. Anomalies: Color-coded alerts from AI/ML agent.
6. Settings: Org/sensor preferences.
7. Monitoring: Admin access to Grafana.

1. Temperature Sensor
2. Humidity Sensor
3. Pressure Sensor
4. Air Quality (CO₂)
5. Motion/Infrared Sensor
6. pH Sensor (chemical)
7. Sound Level Meter
8. Light Sensor (LDR)
9. Vibration Sensor
10. Voltage Sensor

Each sensor has unique edge cases and forecasting patterns.

MIT License. Open for academic and commercial contributions.