Factory Data Analytics Flashcards
7 cards from real SACA practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 7 Factory Data Analytics flashcards as text
A manufacturer collects vibration data from 500 sensors every second. Which storage architecture best handles this volume?
Answer: Time-series database optimized for high-frequency writes
Time-series databases like InfluxDB are architected specifically for high-frequency, timestamped data ingestion from many sources.
What does a scatter plot between tool wear and surface roughness primarily reveal?
Answer: The correlation or relationship between the two variables
A scatter plot visualizes the relationship or correlation between two continuous variables.
In predictive maintenance, what is 'remaining useful life' (RUL)?
Answer: The estimated time until a component will require replacement or failure
RUL is the predicted time or operational cycles left before a component fails, enabling proactive maintenance scheduling.
Which statistical test would you use to determine if defect rates differ significantly across three production lines?
Answer: Chi-square test or ANOVA
Chi-square tests compare categorical outcomes across groups, and ANOVA compares means across three or more groups.
What is feature engineering in the context of factory machine learning models?
Answer: Creating derived input variables from raw data to improve model performance
Feature engineering transforms raw sensor data into meaningful derived variables, such as rolling averages or rate-of-change, that better represent process behavior.
A factory dashboard refreshes every 30 seconds with live OEE data. This is classified as:
Answer: Near real-time analytics
Near real-time analytics involves very short latency (seconds to minutes) between data collection and display, contrasted with true real-time (sub-second) or batch processing.
Which concept describes the practice of training a predictive model on one production line and applying it to a new line with minimal retraining?
Answer: Transfer learning
Transfer learning reuses knowledge from a pre-trained model and fine-tunes it on a new but related domain, reducing the data and compute needed.