Bearing diagnostics
The pump sped up. Then the alarm went off.
Do you bring the inspection forward, or take another reading at the new speed? For the analyst on call, a healthy machine at a new speed can look enough like a damaged one to warrant a closer look.
VesselSeed is exploring a second opinion for that decision. The aim is to take the recording an analyst already has, account for shaft speed and load, and show what evidence still points to damage. Sometimes the most useful answer would be which measurement to take next.
See what the first test found
First accelerometer channel, centered, one shared amplitude scale. Adapted from Schnur et al. under CC BY 4.0.
Findings · First bearing study
At the same detection rate, switching phase on saved zero false alarms.
Damage on a bearing’s inner ring can produce a pulse each time a roller crosses it. A conventional check looks for vibration energy around the expected repetition rate as speed changes. We asked whether the timing of those vibrations across short intervals could help an analyst dismiss healthy readings.
| Analysis | Faults detected | Healthy recordings flagged |
|---|---|---|
| Speed-adjusted baseline | 44 / 48 | 47 / 48 |
| Baseline with phase alignment | 44 / 48 | 46 / 48 |
| Same fitted model, phase switched off | 44 / 48 | 46 / 48 |
Each method missed four damaged recordings. Nearly every healthy recording triggered an alarm. The one-alarm improvement over the standalone baseline persisted with phase switched off. On this evidence, we would keep the analyst’s existing workflow.
Per-recording scores (CSV)Reproduce the study (ZIP)
Study design, thresholds and limits
We used 96 two-second recordings from three physical NU206 bearings, before and after artificial inner-ring damage. The subset covers four speed settings, two load settings, one bearing position and two runs. For each test recording, training excluded both that bearing and that speed setting. The dataset samples different speed settings. Changes in speed during a single field alarm remain an open question.
Each method received the same three-axis acceleration at 20 kHz and the measured shaft RPM. The baseline uses envelope power near the first three harmonics of the expected inner-ring defect frequency. The second model adds a phase-alignment score over eight intervals. For the phase-off row, we kept that model’s weights fixed and replaced the score with its training mean.
We matched detection rates retrospectively using held-out scores. With thresholds set using training data alone, the baseline and phase-on models each detected 36 of 48 damaged recordings and flagged 36 and 34 healthy recordings respectively. With phase off, the model detected 27 and flagged 36. With repeated recordings from just three bearings, this secondary result needs an independent test.
Dataset and rig by Christopher Schnur and colleagues at Saarland University. Source recordings · Experiment design
See how speed changes the signal.
In this synthetic signal, increasing shaft speed brings the pulses closer together. Keep speed fixed and increase pulse jitter to see what irregular timing does. Turn the defect impulse down to zero to see shaft vibration and noise alone.
Synthetic vibration
Illustration only. The simulated defect repeats 7.76 times per revolution. The measured results above come from the physical bearings.
Research behind VesselSeed
What did we take from the invention?
Devin Bostick’s U.S. application Data Processing via Prime Coherence describes a method for finding patterns in a waveform. It stores frequency and phase information in a grid, updates each entry using its neighbors, and selects peaks using chirality, phase and harmonic relationships. The USPTO issued a notice of allowance for the amended claims on September 9, 2026.
For the bearing experiment we borrowed one part, the phase-alignment score in equations 4 and 5, and applied it near frequencies where a defect should repeat. The full grid method has yet to be evaluated on this task. Whether the additional steps help distinguish healthy machines is open.
Proposed field trial
Do you review alarms on variable-speed machines?
We’re looking for a vibration analyst with past pump or motor alarms and follow-up inspection records. An anonymized description of one difficult case is enough for a first conversation.
- You keep the inspection outcomes. We agree on a batch of recordings, the available speed and load information, and the fault-detection target before seeing the answers.
- We make the calls independently. We compare VesselSeed with your usual speed-adjusted review on held-out machines. For each case, we record the evidence for inspection or the specific missing measurement.
- Then we open the records. We count false alarms at the same detection rate, missed faults, cases left unresolved and review time. Fewer false alarms at that detection rate would justify a second batch.
Devin Bostick · devin.bostick@paracosmventures.com