Skip to content

MICKAI^® Training AcademyLEARN · EXPERIMENT · VERIFY

Exam · Predictive maintenance AI with accountable human handover

Take the exam and earn your certificate.

A real, server-scored exam for this course. Score strictly above 80% to pass and get a genuine, verifiable certificate.

Written practice, ungraded

Write your response in your own notes, then compare it with the sample answer. This practice is separate from the scored exam and does not affect your certificate.

Practice 1

Use policy.json, the source definitions and the bad proposal above. Produce a five-line evaluation charter. Define the three-day target with both endpoints. Explain what score 80 can and cannot mean. Separate draft evaluation, human review and equipment authority. List two source assumptions the program cannot authenticate.

Compare with a sample answer

The charter evaluates supplied scores against same-asset events with 0 < event day minus anchor day <= 3. Score 80 is ordinal only. The output is a reproducible desk-study report and draft request for responsible-person review. No diagnosis, maintenance change or closure follows. Event completeness and epoch authenticity remain source-owner assumptions.

Ungraded practice. Compare each requested claim, calculation and limit with the worked guide; a completion mark does not verify understanding.

Practice 2

Use readings.csv and policy.json. Work on a copy; the paper route uses the five-row comparison table above. Calculate P1 anchor 16 and explain why anchor 15 is excluded. List all seven held asset/day pairs. Derive nominal and eligible counts without treating held windows as negatives. Explain why the stale P2 row affects three windows.

Compare with a sample answer

P1/16 uses24,25,26 and mean25 within epoch 1; P1/15 crosses epochs. Held pairs are P1/15, P2/20,21,22 and P3/16,17,18. There are 39 nominal windows and 32 eligible, so coverage is 82.05 percent. Every three-day window containing the stale row is held. The program reports one primary reason but the reviewer retains other visible defects.

Ungraded practice. Compare each requested claim, calculation and limit with the worked guide; a completion mark does not verify understanding.

Practice 3

Use the policy and target definition. The optional test suite includes deliberate chronology mutations. Explain the latest permitted development anchor using both delays. Reject or accept development anchor 11 and evaluation anchor 28 with arithmetic. Describe why a five-row gap is not necessarily five days. Name one preprocessing change that would leak future information.

Compare with a sample answer

Development anchor 9 is latest because9+3+2=14. Anchor 11 needs availability16 and is rejected. Evaluation anchor 28 needs outcomes 31 and availability33, outside the supplied horizon30/freeze32. A row gap changes meaning when assets are interleaved or readings missing. A centred average using tomorrow, or a fitted transform estimated on evaluation data, would leak.

Ungraded practice. Compare each requested claim, calculation and limit with the worked guide; a completion mark does not verify understanding.

Practice 4

Use the baseline report or reproduce it with evaluate.py. Compare threshold 70 with 90 on a copy. Calculate the threshold 70 matrix and denominator checks. Explain how70% window recall coexists with 100% event coverage. Reconstruct the seven episodes and three false episodes. Compare threshold 90 event coverage and lead time without claiming a newly validated policy.

Compare with a sample answer

At 70 the matrix is 7,4,3,18 over32 eligible windows; precision7/11 and recall 7/10. The four events are each covered by at least one alert despite three missed positive windows. Eleven alert rows group to seven episodes, three false. At 90 the matrix is 3,1,7,21; E3 is missed, event coverage 75%, and detected warning times3,2,2. Evaluation-informed threshold selection needs fresh validation.

Ungraded practice. Compare each requested claim, calculation and limit with the worked guide; a completion mark does not verify understanding.

Practice 5

Use handover.json, P2 source days 17-19 and the paper contract above. Write a source-linked prospective draft without using E3 as known evidence. Mark the three unsupported action/diagnosis sentences in the table. Calculate total review demand and the two-episode gap. Run or read test 32 and explain why its structural success is insufficient.

Compare with a sample answer

Keep the known score 80, threshold 70, source days 17-19, epoch 0 and mean38. E3 becomes known only on24 and belongs to later evaluation. Remove diagnosis, deferral and completion claims. Seven episodes require126 minutes;90 fits five, leaving two and 36 minutes. Test 32 deliberately passes structural fields despite false free text, so human semantic review is necessary.

Ungraded practice. Compare each requested claim, calculation and limit with the worked guide; a completion mark does not verify understanding.

Practice 6

As the fictional review lead, assemble a portable evidence pack using only the supplied files. A second learner should be able to reproduce and challenge it. State the target, information freeze and separate human-authority boundary. Include the seven holds, window matrix, episode list, event warning times and workload calculation. Include the repaired prospective P2 day 19 handover and its missing-evidence note. Show the P3 outage comparison without removing E4. Record what the code checks, what it deliberately cannot check and the evidence needed for another bounded trial.

Compare with a sample answer

A defensible packet preserves the original89 rows and four labels, records39 proposed/32 eligible windows, matrix7/4/3/18, seven episodes with three false, event leads3/3/3/2 and 126-minute review demand with two episodes pending. It proves the day 14 freeze with development anchors 3-9 and evaluation15-27 plus labels available 32. P3 removal gives22 eligible, matrix6/3/1/12 and only 3/4 event coverage despite85.71% window recall. The handover remains a source-linked draft with no diagnosis or maintenance change. Limitations include event completeness, authentic resets, reviewer authority, free-text truth, causal benefit and performance outside this invented packet.

Ungraded practice. Compare each requested claim, calculation and limit with the worked guide; a completion mark does not verify understanding.

About this exam

Is this exam graded?

Yes. Every question is scored on our server, never in your browser, and you need a score strictly above 80% to pass. The questions are a real, reviewed bank for this course, drawn fresh for each attempt.

Is it only multiple choice?

Yes. This is a multiple-choice knowledge assessment, with at least one question from each chapter in every attempt. Written exercises are separate practice and are not automatically scored. Passing does not establish practical professional competence or independently verify identity.

Is there a time limit?

Yes, 45 minutes, shown as a countdown once you start. If time runs out, whatever you have answered is submitted automatically.

Do I need to download the PDF first?

No. This course's full text is on its own page, so you can read it online and come straight here when you feel ready.

What do you do with my name and email?

We privately record your name, email, request IP, course and privacy acknowledgement for this request. Newsletter consent is separate and optional. See /privacy.html for storage and retention. A passed exam creates a server-stored named certificate. An email confirmation is not identity assurance.

Can I retake it if I fail?

Yes, straight away. Each attempt draws a fresh set of questions and shuffles the answer order, so it will not be identical to your last attempt, and an attempt can only be submitted once.

Is the certificate real?

It is a genuine, server-issued certificate with its own verification page and a unique ID, not a self-declared one. It records that you passed this course's exam on this date. It is not an accredited qualification, a professional licence or proof of employment.