Skip to content

README.md

Latest commit

 

History

History
78 lines (58 loc) · 3.37 KB

README.md

File metadata and controls

78 lines (58 loc) · 3.37 KB

Analytics Microservices (Flask)

Analytics endpoints for time-series sensor payloads used by the Rasa Action Server and other clients.

Run

  • With Docker (recommended): service microservices in a building compose file (e.g. docker-compose.bldg1.yml or docker-compose.bldg2.yml).
    • Typical host: http://localhost:6001 (host → container 6000)
    • Alternate (some bldg3 variants or isolation strategy): http://localhost:6002
    • Health: GET /health
    • Runner: POST /analytics/run
    • See root README.md for port matrix and isolation guidance.

API

  • POST /analytics/run
    • Body:
      • analysis_type (string, required)
      • Sensor data: either flat or nested payload
        • Flat: { sensorName: [ {timestamp|datetime, reading_value}, ... ] }
        • Nested: { groupId: { innerKey: { timeseries_data: [ {datetime, reading_value}, ... ] } } }
      • Optional params per analysis (e.g., acceptable_range, thresholds, method)
    • Returns: { analysis_type, timestamp, results }

Example request (nested):

{
  "analysis_type": "analyze_temperatures",
  "1": {
    "Air_Temperature_Sensor": {
      "timeseries_data": [
        {"datetime": "2025-02-10 05:31:59", "reading_value": 22.5},
        {"datetime": "2025-02-10 05:33:00", "reading_value": 23.0}
      ]
    }
  }
}

Available analyses (selection)

  • Environmental: analyze_temperatures, analyze_humidity, analyze_co2_levels, analyze_pm_levels, analyze_noise_levels, analyze_formaldehyde_levels, analyze_air_quality, compute_air_quality_index
  • HVAC/air: analyze_supply_return_temp_difference, analyze_air_flow_variation, analyze_pressure_trend, analyze_air_quality_trends, analyze_hvac_anomalies
  • Generic: correlate_sensors, aggregate_sensor_data, analyze_sensor_trend, detect_anomalies, detect_potential_failures, forecast_downtimes
  • Ops: analyze_device_deviation, analyze_failure_trends, analyze_recalibration_frequency

Notes:

  • All functions accept flat or nested payloads and normalize timestamps.
  • Results include human units and UK indoor defaults (where applicable): °C, %RH, ppm, µg/m³, etc.
  • Anomaly detection supports method: zscore or iqr (robust).

UK defaults and units

Embedded UK-oriented guidelines include:

  • Temperature: 18–24 °C; Humidity: 40–60 %RH
  • CO2: good 400–1000 ppm; max 1500 ppm
  • PM2.5: 35 µg/m³; PM10: 50 µg/m³
  • Formaldehyde (HCHO): 0.1 mg/m³; Noise: ~55 dB(A)

You can override thresholds/ranges in the request. Responses include unit, acceptable_range, or acceptable_max metadata where relevant.

Integration

  • Action Server calls this service via internal URL http://microservices:6000/analytics/run (see rasa-ui/actions).
  • Decider Service optionally selects which analysis_type to run for a user question.

Customize for your building

  • Sensor naming: key detection is robust (e.g., matches "temperature" and "temp" while avoiding "attempt"). Prefer human-readable names in payloads.
  • Thresholds: pass acceptable_range or thresholds per request to tailor to building standards.
  • New analyses: add a new function in blueprints/analytics_service.py and register it in the analysis registry used by /analytics/run.

Testing

  • Health: open http://localhost:6001/health
  • Smoke test: python microservices/test_analytics_smoke.py — exercises all registered analyses with dummy data and prints a summary.