🛰️

Catenary Corrosion Intelligence System

⬡ Track 1 — Multispectral Input Parameters
⬡ Track 1 — Catenary Corridor Corrosion Severity Map
⬡ Track 1 — LightGBM Classification Output
Multispectral Sensor Feed
82

Corridor Severity Profile — Railway Line Survey
Class 0–1 Safe Class 2–3 Caution Class 4 Critical
6-mo Forecast
—
Predicted class
Degradation Rate
—
Class/year
LGBM Confidence
—
Model certainty
Next Inspection
—
Recommended
LightGBM Classification GBDT
--
AWAITING ANALYSIS
Run classification to begin
Confidence: —
Corrosion Class —
Iron Oxide Ratio (R/B) —
Norm. Rust Index —
NIR Suppression —
Maintenance Action —
NDVI
—
NDRE
—
GNDVI
—
OSAVI
—
LCI
—
IOR
—
AI Recommendation
System ready. Input 5-band reflectance values and run classification.
LightGBM Predictive Engine — Feature Importance & Time-Series Degradation Forecast
Algorithm
LightGBM
Gradient Boosted
Trees
500
Estimators
Max Leaves
63
Leaf-wise
F1 Score
0.924
5-class OvA
IOR
0.92
NRI
0.74
NDRE
0.61
NIR-S
0.55
Humidity
0.38
Temp
0.22
Current State
—
—
3-Month Projection
—
—
6-Month Projection
—
—
12-Month Risk
—
—
⬡ Track 2 — LiDAR Infrastructure Geometry & Terrain Monitoring (Zenmuse L2)
Catenary Sag
342 mm
Within tolerance (±50mm)
Pole Verticality
1.8°
Deviation within limit
Embankment Slope Δ
+12 mm
Monitor — monsoon season
Clearance Envelope
6.42 m
Above 5.6m minimum
🤖 AI Corrosion Engineer ✕
Hello! I'm your AI Corrosion Engineer. Ask me about LightGBM classification, spectral indices, corrosion severity, LiDAR monitoring, or predictive maintenance strategy.