Math Bridge: Why can't aggregation repair aliasing?

← Back to Edge Processing Patterns
Math BridgeAnalytics & MLStruggle-friendly runway

Why can't aggregation repair aliasing?

Fold one motor tone first, then measure what four summary values really save.

Data Dora, the guideData Dora guides
The one targetShow why a 650 Hz tone becomes a believable 350 Hz feature.
The chapter case1 kHz vibration; 1 Hz temperature; 3,600 values become four.
What it buys youBandwidth savings without pretending aggregation validates inputs.

A field team faces an unresolved physical question: Why can't aggregation repair aliasing? They must answer it before changing real tone on the real device. Predict the direction first.

See the relationship before changing it

The figure reads from left to right. The blue card is real tone. The middle card applies this page's relationship. The green card is temperature ceiling. Walk the arrows once: set the input, apply the rule, then read the result with its unit.

The retained audit below checks several chapter fixtures. This added model holds every other chapter fixture fixed, so the numeric fixture does not switch without explanation.

Real tone changes temperature ceiling An input card leads through the page relationship to the temperature ceiling result. SET INPUT ONE CONTROL APPLY RULE predict calculate check units READ RESULT
Walk the arrows. The sample grid decides the apparent frequency before the aggregation formula runs. Summary statistics then compress that already-aliased sequence.

Derive the baseline in four named moves

  1. 1

    Name the input. The chapter baseline for real tone is 650.

  2. 2

    Name the relationship. 1 Hz → 0.5 Hz ceiling; 1,000 Hz → 500 Hz ceiling falias=|650-1(1,000)|=350 Hz 3,600/4=900x; (1-4/3,600)100=99.89%

  3. 3

    Substitute the chapter fixture. Set real tone to 650. The page ledger gives temperature ceiling as 0.5 Hz.

  4. 4

    Read the result. Keep Hz beside the value. Use it only inside the technical boundary on this page.

Predict, then change real tone

Try Predict the direction of temperature ceiling. Move one control, calculate, then check your prediction.

650
Chapter baseline
Temperature ceiling

Observe The sample grid decides the apparent frequency before the aggregation formula runs. Summary statistics then compress that already-aliased sequence. Reset the control to 650 and compare temperature ceiling.

Explain Only real tone moves here. The other chapter fixtures remain fixed.

Check yourself

What should you do before trusting a moved-control result?
Answer: Predict its direction, apply the shown relationship, keep the units, and reset to the worked baseline.
What does this small model leave out?
Answer: Only real tone moves. Field effects named in the page's technical boundary stay fixed.

Honesty note. This teaching model changes only real tone. It replays the chapter fixture; it is not a field prediction.

1. Ask whether each sample is real

Minimum, maximum, mean, and standard deviation faithfully describe the samples they receive. They cannot tell whether those samples already contain a folded frequency.

Data Dora: A good summary of bad samples is still bad evidence.

2. Name every algebra move

1

Halve each rateThe Nyquist ceiling is fs/2.

2

Choose the nearest spectral copyk=round(fsignal/fs).

3

Measure the foldfalias=|fsignal−kfs|.

4

Divide raw by summaryCompression=3,600/4.

5

Convert retained share to savingReduction=(1−4/3,600)×100%.

3. Reproduce the chapter case

1 Hz → 0.5 Hz ceiling; 1,000 Hz → 500 Hz ceiling
falias=|650−1(1,000)|=350 Hz
3,600/4=900×; (1−4/3,600)100=99.89%

The 350 Hz result fits inside the allowed band, so the aggregator has no mathematical clue that it came from 650 Hz.

4. Try the vibration tone

TryMove the real tone across the 500 Hz ceiling and watch the reported frequency turn back toward zero.

Real tone
Temperature ceiling
Vibration ceiling
Spectral copy k
Apparent tone
Raw values
Summary values
Compression
Reduction

ObserveAt 650 Hz the output is 350 Hz. At 750 Hz it is 250 Hz. The 900.00× compression and 99.89% saving remain unchanged.

ExplainThe sample grid decides the apparent frequency before the aggregation formula runs. Summary statistics then compress that already-aliased sequence.

Technical boundaries.

This compact engine tracks one steady tone and value-count compression.

Signal
Real motor spectra contain harmonics, modulation, noise, and transients
Bandwidth
Bytes, headers, timestamps, retries, and encoding can change network savings
Statistics
Four values do not preserve phase, waveform shape, rare events, or provenance

Validate sensor bandwidth, filtering, window semantics, and the decision that consumes each summary.

5. Choose the boundary before the pattern

Filter before sampling, confirm the retained band, then choose a window and statistics that preserve the decision-relevant feature.

6. Keep the aggregation record

Record sensor and sample bands, anti-alias filter, window length and alignment, raw count, summary fields, units, missing-data rule, downstream threshold, owner, and retest trigger.

7. Check yourself

Why is 650 Hz unsafe on a 1 kHz grid?
Answer: The safe band stops at 500 Hz.
Why does it appear at 350 Hz?
Answer: The nearest copied spectrum is at 1,000 Hz, so |650−1,000|=350 Hz.
Does a 99.89% reduction prove a valid feature?
Answer: No. It proves a value-count saving, not that the input samples represented the real signal.
Honesty boundary.

The sample and aggregation values come from the chapter; the bearing tone is labelled illustrative.

1 Hz and 1 kHz
The chapter's temperature and vibration rates
3,600 to four
The chapter's hourly aggregate example
650 Hz
An illustrative plausible bearing harmonic, not a measured machine value

Correct, not complete: the Aggregate pattern does not qualify the acquisition chain or anomaly model.

Return to Edge Processing Patterns with the maths unpacked →