SACA Smart Factory Data Analytics 4 — Questions and Answers
Question 1: Which protocol is most commonly used to transport IoT sensor data from factory floor devices to a broker in a smart factory?
- HTTP/1.1
- MQTT (Correct answer)
- FTP
- SMTP
Correct answer: MQTT
MQTT is a lightweight publish-subscribe protocol designed for constrained IoT devices and low-bandwidth, high-latency networks typical in factory environments.
Question 2: A factory analyst wants to identify which combination of machine parameters most frequently precedes a product defect. Which analytical approach is most appropriate?
- Descriptive statistics on defect count alone
- Association rule mining on process parameter logs (Correct answer)
- Linear regression on throughput data
- Network topology mapping
Correct answer: Association rule mining on process parameter logs
Association rule mining discovers frequent co-occurring parameter combinations in event logs, revealing multi-factor patterns that precede defects.
Question 3: What is 'data drift' in the context of a smart factory ML model deployed in production?
- Data being physically moved between servers
- A gradual change in the statistical distribution of incoming data compared to training data (Correct answer)
- Loss of data due to network packet drops
- Incorrect timestamp alignment across sensors
Correct answer: A gradual change in the statistical distribution of incoming data compared to training data
Data drift occurs when the real-world input distribution shifts over time, causing a model trained on older data to lose prediction accuracy.
Question 4: In a smart factory dashboard, a line chart shows sudden spikes in cycle time every Monday morning. What is the most likely root cause to investigate first?
- Random sensor noise
- Network bandwidth saturation in the data warehouse
- A weekly scheduled maintenance or shift changeover event (Correct answer)
- A bug in the dashboard rendering library
Correct answer: A weekly scheduled maintenance or shift changeover event
Periodic spikes aligned with a calendar pattern (Monday mornings) strongly suggest a recurring scheduled event such as maintenance, startup, or shift handover.
Question 5: Which technique reduces the dimensionality of high-dimensional sensor data while preserving the most variance, aiding visualization and model training?
- K-means clustering
- Principal Component Analysis (PCA) (Correct answer)
- Random Forest feature selection
- Long Short-Term Memory (LSTM) networks
Correct answer: Principal Component Analysis (PCA)
PCA projects high-dimensional data onto a lower-dimensional space that captures the maximum variance, making it easier to visualize patterns and train models.
Question 6: A smart factory collects temperature readings from 500 machines every second. After 30 days, the raw dataset is approximately how large if each reading is a 4-byte float?
- ~52 MB
- ~5.2 GB (Correct answer)
- ~52 GB
- ~520 GB
Correct answer: ~5.2 GB
500 machines × 1 reading/sec × 86,400 sec/day × 30 days × 4 bytes ≈ 5.18 GB, illustrating the data volume challenges in industrial IoT.
Question 7: What is the role of a message broker like Apache Kafka in a smart factory data pipeline?
- It stores long-term historical analytics results
- It acts as a durable, high-throughput buffer between data producers and consumers (Correct answer)
- It performs SQL queries on streaming data
- It manages machine PLC programs
Correct answer: It acts as a durable, high-throughput buffer between data producers and consumers
Kafka decouples producers (sensors, PLCs) from consumers (analytics engines, databases) by buffering messages durably at high throughput.
SACA Smart Factory Data Analytics 4
Which protocol is most commonly used to transport IoT sensor data from factory floor devices to a broker in a smart factory?