SACA Factory Data Analytics 5 — Questions and Answers
Question 1: What is the primary risk of using only accuracy as the evaluation metric for a factory defect detection model when defects are rare?
- Accuracy cannot be calculated for imbalanced datasets
- A model predicting 'no defect' always will achieve high accuracy but miss all defects (Correct answer)
- Accuracy overestimates false positives in rare-event scenarios
- Accuracy is undefined when classes are unbalanced
Correct answer: A model predicting 'no defect' always will achieve high accuracy but miss all defects
With imbalanced classes, a naive model that never predicts a defect can still achieve 99%+ accuracy while providing zero practical value.
Question 2: In factory analytics, what is a digital twin primarily used for?
- Storing backup copies of production databases
- Creating a virtual simulation of a physical asset for analysis and optimization (Correct answer)
- Duplicating sensor data across redundant servers
- Generating synthetic data to augment small training sets
Correct answer: Creating a virtual simulation of a physical asset for analysis and optimization
A digital twin is a virtual model of a physical asset that mirrors real-time behavior, enabling what-if analysis and predictive optimization.
Question 3: Which edge computing benefit is most critical for real-time quality inspection on a high-speed production line?
- Lower storage costs compared to cloud databases
- Reduced latency by processing data locally near the source (Correct answer)
- Easier integration with ERP systems
- Simplified model training without cloud resources
Correct answer: Reduced latency by processing data locally near the source
Edge computing processes data at or near the source, eliminating round-trip cloud latency that would be unacceptable for millisecond-level inspection decisions.
Question 4: A process engineer wants to understand which input variables most influence yield. Which analytical technique is most appropriate?
- Time-series decomposition
- Design of Experiments (DoE) with main effects analysis (Correct answer)
- Run chart with seasonal adjustment
- K-nearest neighbor classification
Correct answer: Design of Experiments (DoE) with main effects analysis
Design of Experiments systematically varies input factors to measure their individual and interaction effects on a response variable like yield.
Question 5: What does the 'Performance' component of OEE measure?
- The percentage of scheduled time the machine was running
- The ratio of actual production speed to the theoretical maximum speed (Correct answer)
- The proportion of produced parts that meet quality standards
- The mean time between maintenance interventions
Correct answer: The ratio of actual production speed to the theoretical maximum speed
OEE Performance measures how fast the machine runs compared to its ideal speed, capturing slow cycles and minor stops.
Question 6: Which data governance practice ensures that factory sensor data can be traced back to its source, collection time, and transformations applied?
- Data masking
- Data lineage tracking (Correct answer)
- Data deduplication
- Schema validation
Correct answer: Data lineage tracking
Data lineage documents the origin, movement, and transformation of data throughout its lifecycle, supporting auditability and troubleshooting.
Question 7: A factory implements a rules-based alert that triggers when pressure exceeds 150 PSI. This is an example of which analytics type?
- Predictive analytics
- Prescriptive analytics
- Threshold-based descriptive monitoring (Correct answer)
- Diagnostic root cause analysis
Correct answer: Threshold-based descriptive monitoring
A fixed threshold alert describes current conditions and flags deviations in real time without predicting future states or recommending actions.
SACA Factory Data Analytics 5
What is the primary risk of using only accuracy as the evaluation metric for a factory defect detection model when defects are rare?