Skip to main content
Version: Next

BI Analysis and Diagnosis

1. From reporting to industrial analysis

Prodia does not stop at displaying charts.

It organizes manufacturing analytics around operational questions such as output, quality, takt, fault, SPC, and OEE, then continues into diagnostic reasoning and investigation guidance.

2. Typical analytical domains

DomainTypical focus
Outputline output, process throughput, shift comparison, trend
Qualitydefect distribution, bad rate, FPY, quality loss
Takt and balancebottleneck process, station fluctuation, line balance
Fault and downtimedowntime duration, Pareto, MTBF, MTTR
OEEavailability, performance, quality decomposition
Process parametersSPC, parameter drift, abnormal window, relationship analysis

3. Analytical depth

Prodia commonly moves across four layers:

LayerCore questionTypical methods
DescriptiveWhat happened?query, aggregation, ranking, distribution
DiagnosticWhy did it happen?comparison, correlation, Pareto, root-cause clues
PredictiveWhat is likely to happen?risk signals, deterioration trend, threshold judgment
PrescriptiveWhat should be done?next checks, operational suggestion, optimization path

4. Typical output structure

When Prodia performs BI analysis and diagnosis, it usually returns a combination of:

  • key metrics
  • comparison results
  • anomaly concentration points
  • likely factor ranking
  • recommended next checks

This helps users move directly from seeing a number to deciding what to investigate next.

5. Why this differs from general BI

General BI tools often require users to:

  • know where the data lives
  • know which chart to build
  • know how to interpret the result
  • know how to continue into diagnosis

Prodia compresses this path by making analytics queryable through industrial language and by keeping diagnosis linked to the same result context.

6. Best-fit problems

This capability is strongest when teams need to answer questions such as:

  • which object changed the most
  • where the main loss comes from
  • which abnormal factor deserves first attention
  • how one indicator differs across line, process, station, shift, or period