Skip to main content

​
GET /api/v1/report/?format=industrial

Generate a comprehensive 5-page Industrial Asset Health Certificate for engineers and maintenance teams. This is the recommended report format for technical audiences.

​
Path Parameters

​
asset_id
string
required
Asset identifier (e.g., Motor-01)

​
Query Parameters

​
format
string
required
Must be industrial

​
Response

Returns a professional PDF file with:
  • 5 pages of detailed technical analysis
  • ML explainability with feature contribution breakdown
  • Maintenance correlation with operator log events
  • Sensor trend charts with anomaly annotations
  • Audit trail with model version and data provenance

​
Example

cURL
curl -X GET "https://predictive-maintenance-uhlb.onrender.com/api/v1/report/Motor-01?format=industrial" \
  -H "Accept: application/pdf" \
  --output motor01-industrial-cert.pdf
Python
import requests

response = requests.get(
    "https://predictive-maintenance-uhlb.onrender.com/api/v1/report/Motor-01",
    params={"format": "industrial"}
)

with open("motor01-industrial-cert.pdf", "wb") as f:
    f.write(response.content)

​
File Naming

Downloaded files follow the pattern: 
industrial-health-cert_{asset_id}_{timestamp}.pdf
Example: industrial-health-cert_Motor-01_2026-03-02T10-30-00Z.pdf

​
Report Structure

​
Page 1: Executive Summary

  • Header: Company branding, asset ID, report date
  • Health Status Card: Grade (A-F), Score (0-100), Risk Level badge
  • Key Metrics Grid:
    • Degradation Index (DI) with progress bar
    • Remaining Useful Life (RUL) in hours
    • Damage Rate (per hour)
    • Anomaly Detection Statistics
  • Recommendation: Maintenance action required (if any)

​
Page 2: Current Sensor Readings

  • Real-time measurements at report generation time:
    • Voltage (V)
    • Current (A)
    • Power Factor
    • Vibration (g)
    • Computed Power (kW)
  • Baseline comparison: Current vs. Target ranges
  • Deviation analysis: Percentage difference from baseline

​
Page 3: ML Explainability

This page provides transparency into the ML model’s decision-making process.
  • Feature Contribution Breakdown:
    • Batch model features (16-D: mean, std, peak-to-peak, rms for each signal)
    • Legacy model features (6-D: rolling means, spike counts, etc.)
    • Contribution scores showing which features triggered anomaly detection
  • Detection Logic:
    • Anomaly score calculation method
    • Dead-zone threshold (HEALTHY_FLOOR = 0.65)
    • Sensitivity constant (0.005)
  • Model Metadata:
    • Model version (v2 Legacy, v3 Batch)
    • Training date
    • Baseline ID

​
Page 4: Maintenance Correlation Analysis

  • Operator Log Events: Recent maintenance activities from the last 24 hours
  • Correlation Matrix: Time-based correlation between:
    • Operator log timestamps
    • Anomaly detection events
    • Health score degradation
  • Causal Analysis: Identifies whether maintenance events:
    • Preceded anomalies (predictive)
    • Followed anomalies (reactive)
    • Improved health scores (effective)
This analysis helps validate that maintenance activities are having the intended effect on asset health.

​
Page 5: Sensor Trend Charts & Audit Trail

  • Time-Series Charts (last 60 seconds):
    • Voltage trend with baseline range overlay
    • Current trend with spike annotations
    • Vibration trend with anomaly markers
    • Health score timeline with risk level transitions
  • Audit Trail:
    • Report generation timestamp
    • Data source (InfluxDB bucket)
    • ML model versions
    • Operator who requested the report (if authenticated)
    • Hash of sensor data (for tamper detection)

​
Use Cases

Root Cause Analysis

Investigate why anomalies occurred

Maintenance Planning

Schedule preventive maintenance based on RUL

Regulatory Compliance

Provide auditable evidence of asset monitoring

Insurance Claims

Document asset condition at time of failure

​
The Snapshot Rule

All reports use the Snapshot Rule: Sensor data and health metrics are captured at the moment of report generation. This ensures reports are immutable and auditable.
The Industrial Certificate includes a cryptographic hash of the sensor data to prevent tampering. This makes it suitable for:
  • Regulatory audits (FDA, ISO 55000)
  • Insurance claims
  • Legal proceedings
  • Safety investigations

​
Status Codes

  • 200 - Certificate generated successfully
  • 404 - Asset not found or no health data available
  • 503 - PDF generation failed

​
Performance

  • Generation Time: 2-5 seconds (includes chart rendering)
  • File Size: ~500 KB (depends on chart complexity)
  • Charts: Rendered server-side using ReportLab
Generate Industrial Certificates weekly for high-value assets, or on-demand when anomalies are detected. Store certificates in a document management system for historical comparison.

​
Comparison with Other Reports

FeatureExecutive PDFAnalyst ExcelIndustrial PDF
Pages13 sheets5
AudienceExecutivesData AnalystsEngineers
ML Explainability❌ No❌ No✅ Yes
Sensor Charts❌ No❌ No✅ Yes
Operator Logs❌ No✅ Yes✅ Yes
Maintenance Correlation❌ No❌ No✅ Yes
Audit Trail❌ No❌ No✅ Yes
Generation Time1s~2s2-5s

Executive PDF

1-page summary for quick review

Analyst Excel

Raw data for statistical analysis

Build docs developers (and LLMs) love

Get started for freeTalk to us