Factory Analytics | Factory Intelligence Laboratory

Factory Analytics

Factory Analytics develops robust and transferable AI methods for diagnosing faults, predicting degradation, and discovering causal quality relationships across machines, products, processes, and domains.

Research StackFault Detection · Prognostics · Causal Quality

Research Directions

Next-Generation Fault Detection, Prognostics, and Causal Quality Analytics

Feature Engineering & Process Fault Detection

Feature representation, signal analysis, and process fault detection methods for manufacturing monitoring and machine health assessment.

Adaptive Fault Detection and Prognostics

Continual domain adaptation, source-free adaptation, open-set domain generalization, target-free generalization, and diffusion-assisted cross-domain fault diagnosis.

Multi-stage Quality Causality

Causal process-genealogy mining for discovering how upstream process conditions, machine states, material histories, and decisions affect downstream quality.

Event-driven Machine Failure Analysis

Failure diagnosis using event logs and system-state transitions to reason about root causes, functional interactions, and product field failures.

Sound Source Localization

Noise detection and sound-source localization for acoustic quality monitoring in clamorous manufacturing environments.