Dashboard

Fleet overview, uncertainty analysis, engine deep-dive, and SHAP explainability in one place.

Engine-level exploration of predictions, uncertainty, and feature attribution.

FD001 test enginesScope: last window per engine (100 engines)RUL cap: 125 cycles

Decision Support

1) Fleet Overview

Global metrics stay fixed; triage charts and table in this section update with Risk Tier and Min Confidence filters.

RMSE

12.51

Global metric (fleet-wide), lower is better

NASA Score

274.4

Global metric (fleet-wide), penalty-weighted error

Risk Quadrant Scatter (RUL Mid vs Width)

Legend

RUL threshold for monitor (RUL_MID = 40)

Median Uncertainty (50th Quantile of Interval Width = 37.9)

High risk engine

Medium risk engine

Low risk engine

Risk Distribution

Number of engines in respective categories

High

Medium

Low

Filters & Interaction Controls

Engine Selector

Risk Tier

Min Confidence: 0%

05101520253035404550556065707580

Current slider range: 0% to 80%.

Top 20 Priority Engines

Engine	RUL Low	RUL Mid	RUL High	Width	Confidence	Risk Tier
Engine 31	5.1	7.4	9.2	4.1	0.90	High
Engine 68	5.1	7.4	9.2	4.1	0.90	High
Engine 81	5.1	7.5	9.2	4.1	0.90	High
Engine 34	5.4	7.8	9.6	4.2	0.90	High
Engine 76	5.6	8.0	9.9	4.4	0.89	High
Engine 82	7.6	10.4	12.9	5.4	0.87	High
Engine 42	8.2	11.2	14.0	5.7	0.86	High
Engine 35	9.1	12.3	15.4	6.3	0.85	High
Engine 20	11.0	14.6	18.2	7.2	0.83	High
Engine 49	11.0	14.7	18.4	7.4	0.83	High
Engine 56	12.3	16.1	20.2	7.9	0.81	High
Engine 66	13.7	17.9	22.5	8.8	0.80	High
Engine 100	15.8	20.4	25.7	9.9	0.77	High
Engine 24	17.6	22.7	28.6	11.0	0.75	High
Engine 61	17.9	23.0	29.0	11.1	0.75	High
Engine 41	18.7	24.0	30.2	11.5	0.74	High
Engine 77	19.4	24.8	31.2	11.8	0.74	High
Engine 64	19.5	24.9	31.4	12.0	0.73	High
Engine 92	19.5	24.9	31.4	11.9	0.73	High
Engine 36	20.7	26.4	33.3	12.6	0.72	High

2) Prediction Quality & Uncertainty

These quality views are fleet-level and respond to Risk Tier and Min Confidence filters.

Prediction Interval Plot

Coverage Summary

Coverage on filtered fleet: 83.0%

Signed Error Plot (True RUL - Predicted RUL)

3) Engine Deep Dive (Selected Engine)

Panels in this section update when you change the selected engine.

Predicted RUL (median)

Interval [--, --]

True RUL

Ground-truth value for selected engine

Confidence

Width -- cycles

Coverage status

Global coverage 83.0%

Risk Signal (Selected Engine)

No engine selected.

Engine Phase Split and Capping Rationale

Engine series payload is not available for this selection.

Pooled Sensor Degradation (Mean ± Std Across Engines)

Local SHAP (Selected Engine)

No local SHAP data for selected engine.

4) Interpretability (Global)

Global SHAP stays constant across engine and filter selections.

Global SHAP Top Features