Operational Excellence Decision Support

Find the leading indicators most likely to move a target KPI.

This demo loads a manufacturing KPI history file, maps every column into the PCOEx hierarchy, fits the decision model, and turns the result into ranked improvement vectors.

Current dataset Loading KPI history
-Periods
-KPI columns
-Invalid rows

Guided workflow

What The Demo Is Doing

Loading
1

Load KPI History

Reads the dummy CSV file and treats the first column as the time period.

2

Map The Hierarchy

Matches each column to a people/process category, layer, and level.

3

Choose Target KPI

Selects the outcome the team wants to improve or protect.

4

Run The Model

Fits the PLS response model and prepares CCA category validation.

5

Rank Action Vectors

Converts model importance into clear improvement recommendations.

Step 3

Select The Outcome

Pick one of the five PCOEx response variables used in the dissertation case: Total Revenue, Net Sales, Stress Level, Turnover Rate, or Absenteeism Rate.

Export Summary
Recommended first focus

Waiting for model

The highest-ranked action vector will appear here after analysis completes.

Confidence -
Selected target - The outcome the analysis explains.
R2 fit - How much target variation the model explains.
Prediction error - RMSE in the target KPI's unit.
Rows used - Valid periods included in the fit.
Predictors used -
Mean absolute error -
Excluded predictors -

Literature workflow

How The Recommendation Is Calculated

The demo follows the PCOEx-PMS literature visuals: PLS model selection, response plot, residual plot, coefficient table, and canonical-covariate heatmap.

Overall confidence score -
Cross-validated R2 -
Cross-validated RMSE -

PLS Model Selection Plot

PLS Response Plot

PLS Residual Plot

PLS Coefficients

Step 5

Ranked Action Vectors

Each row is an actionable leading indicator. The practical label is paired with the literature-loyal statistical interpretation: coefficient direction and whether operational review is required.

Step 4

Heatmap Canonical Covariates PCOEx-PMS

This follows the dissertation heatmap style for CCA category validation. Each cell is the canonical correlation between a PMEI category covariate and the selected response variable.

Steps 1 and 2

Data Quality Check

The import step confirms the file shape before modeling. Bad values are counted and kept out of the fit.

Periods
-
Mapped columns
-
Observations
-
Invalid observations
-

Configured KPI map

PCOEx Hierarchy

This is the map the model uses to separate actionable leading indicators from lagging outcomes.