Analytics & ML · Study deck
Edge Data: Sampling and Statistical Compression
This first route protects signal evidence through sampling and compares lossless, statistical, and transform reduction.
Data Dora is your guide for this deck.
After studying this chapter
Learning objectives
You will be able to:
- Apply Nyquist Theorem: Calculate appropriate sampling rates for different sensor types
- Implement Data Reduction Techniques: Use aggregation, compression, event-based reporting, and delta encoding
- Select Compression Algorithms: Choose optimal algorithms based on data type and edge device constraints
- Avoid Common Pitfalls: Prevent sampling aliasing, buffer overflow, and rate mismatch errors
Major section
Start With the Story
If the device measures too rarely, no later data trick can rebuild the missing motion.
- Sampling means taking readings at chosen times.
- The sampling rate is how many readings are taken in one second.
- Compression means using fewer bits to carry useful information.
- Exact records may need a lossless method, which can rebuild every bit.
Major section
Start With the Story (continued)
Trends may allow summaries.
- The simple rule has limits.
- Less data saves power and link use, but it can also hide faults.
- A device has produced messy evidence, an analytic step is about to change an alert or control decision, and someone has to explain why the result should be trusted.
Major section
Edge Sampling Compression Basics
Edge sampling and compression reduce the amount of data IoT devices need to transmit.
- By transmitting only important changes or compressed summaries, devices save battery power and network bandwidth while preserving the information that matters most.
Major section
Edge Data Reduction Basics
Core Concept: Transform raw sensor data into actionable information at the source - send summaries, statistics, and alerts rather than every reading.
- A sensor sending 1000 samples/minute to the cloud uses 100x more bandwidth than one sending minute-averages - with identical analytical value for most applications.
Major section
Key Concepts
Adaptive sampling: Dynamically adjusting the sensor sampling rate based on signal variance or event rate — increasing frequency when the signal changes rapidly and decreasing it during quiet periods.
- Delta encoding: A compression technique transmitting only the change between consecutive readings rather than absolute values, highly effective for slowly varying sensors.
Major section
Phoebe's Field Notes: Why Sampling First Protects Compression
Eight rolling elements place the fault at 240 Hz and its third harmonic at 720 Hz, so strict Nyquist needs 1,440 Hz and the chapter's 2.5× margin needs 3,600 Hz.
- At 4 kHz and two bytes per sample, raw data is 8,000 B/s; a 120 B/s peak record is a 66.7× reduction without discarding the input's 16-bit resolution.
Major section
Try It: Edge Aggregation Data Reduction
Adjust the sensor sampling rate and aggregation window size to see how edge aggregation reduces data volume.
- The widget calculates the bandwidth reduction and shows what information is preserved versus lost.
Deck summary
Key takeaways
If the device measures too rarely, no later data trick can rebuild the missing motion.
- Trends may allow summaries.
- Edge sampling and compression reduce the amount of data IoT devices need to transmit.
- Core Concept: Transform raw sensor data into actionable information at the source - send summaries, statistics, and alerts rather than every reading.
- Adaptive sampling: Dynamically adjusting the sensor sampling rate based on signal variance or event rate — increasing frequency when the signal changes rapidly and decreasing it during quiet periods.
Retrieval practice
Recall check
Data Dora says: answer from memory, then check your reasoning.
Q1A motor spins at 60 Hz and bearing defects appear at the 2x, 3x, and 4x harmonics. What minimum sampling rate avoids aliasing these fault signatures?
Show answer
Answer: A Nyquist requires sampling at twice the highest frequency of interest; the 240 Hz 4th harmonic needs 480 Hz.
Print reference
Answers
Answer key.
- A · Nyquist requires sampling at twice the highest frequency of interest; the 240 Hz 4th harmonic needs 480 Hz.