Selbstgesteuertes Projekt
Predictive Maintenance: Industrial ML for Fault Detection
Condition-Monitoring Engineer · KI / ML · Juli 2026
Eine ProoV-Fallstudie · Lernprojekt, kein Arbeitsverhältnis
Built an end-to-end rolling-bearing fault detector using vibration snapshots, physics-based time and frequency features, and a compact decision tree sized for embedded use. Evaluated on held-out bearings with a full confusion matrix, then wrote a maintenance memo that justified a 0.40 alarm threshold using the candidate’s own catch-rate, false-alarm rate, and cost numbers while correctly localizing the fault to the outer race from BPFO.
Bewertet nach
- Feature engineering (signal to honest numbers)25%
- Held-out evaluation (honest measurement)25%
- Cost-based threshold decision + memo25%
- Compact model + embedded-budget justification15%
- Plain-language explanation10%
Bestanden · Bestehensgrenze 60/100
Was herausstach6
- Computed RMS, kurtosis, and crest factor with step-by-step logic and showed the expected spike in kurtosis/crest factor on the faulty signal.
- Used envelope-spectrum energy at the supplied defect frequencies and correctly found BPFO as the dominant band, mapping it to the outer race.
- Trained a depth-capped decision tree on training bearings only and stated the tree depth and node count against an embedded budget.
- Reported a complete held-out confusion breakdown with catch-rate and false-alarm rate, then defended a threshold using explicit cost figures.
- Leakage-Free Feature Engineering
- Honest Held-Out Evaluation
Meine eingereichte Arbeit18 Aufgaben
10 Aufgaben · 11.5k Zeichen · Python · 8 Textantworten · 714 Wörter
- Initial Prediction62 Zeichen
- Load Plot Signal CodePython · 1.8k Zeichen
- Example Rms WorkedPython · 697 Zeichen
- Time FeaturesPython · 1.9k Zeichen
- Freq FeaturesPython · 1.1k Zeichen
- Fault Location239 Zeichen
- Train Classifier363 Zeichen
- Evaluation HeldoutPython · 3.8k Zeichen
- Threshold Decision334 Zeichen
- Decision Memo1.2k Zeichen
- Initial Prediction9 Wörter
- Load Plot Signal Code163 Wörter
- Example Rms Worked72 Wörter
- Time Features173 Wörter
- Freq Features110 Wörter
- Fault Location14 Wörter
- Train Classifier31 Wörter
- Evaluation Heldout142 Wörter
Bewusst zusammengefasst — der eingereichte Code und die Texte bleiben privat, damit diese Seite nicht als Lösungsvorlage dient. Die vollständige Einreichung steht hinter dem verifizierten Zertifikat.