# Autonatomy v269 — Classroom Pilot Observation Protocol

## Purpose
Use the Electrical pilot to learn where the product helps or gets in the learner's way before more architecture or course modernization is added.

## Pilot scope
Recommended: 5–8 learners, 45–60 minutes, instructor-led access to `tools/electrical-pilot-journey.html`.

The six checkpoints remain unchanged from v268:
1. Electrical quantities
2. Circuit analysis
3. Test instruments / circuit diagnosis
4. Starting & charging diagnosis
5. Evidence Portfolio
6. Practical Handoff

## What v269 records automatically
All data is browser-local.
- pilot session ID
- page and pilot-step visits
- elapsed time between entering and leaving pilot steps
- repeat visits / backtracking
- learner `I'm stuck` friction markers
- v268 pilot feedback submission

v269 does **not** capture:
- names or identities unless an instructor puts them manually in a note
- keystrokes
- lesson answers
- typed diagnostic responses
- audio, video, screen recording, or camera data

## What the instructor should observe
Do not coach immediately unless safety, access, or a true blocker requires it. Watch for:
- Orientation — does the learner know where they are and what comes next?
- Navigation — do they hunt for buttons, backtrack, or leave the path accidentally?
- Reasoning — do interactions change or clarify their mental model?
- Tool selection — do they choose an instrument/test boundary intentionally?
- Evidence — can they explain what a result proves and what remains unproven?
- Transfer — can they explain what they would verify on a real vehicle?

Use `tools/pilot-observation-center.html` to add notes while observing.

## Suggested observation language
Record behavior rather than judgment.

Better: `Step 3: learner hovered between DMM and test-light options for ~20 sec, then explained voltage drop before choosing DMM.`

Avoid: `Student is weak at meters.`

## After the session
1. Ask the learner to complete the existing clarity/usefulness/friction feedback on the pilot page.
2. Open Pilot Observation Center.
3. Review step timing and friction markers together with instructor notes.
4. Export JSON for complete structured data and/or CSV for spreadsheet review.
5. Do not interpret long time automatically as failure; it may indicate productive reasoning.

## Pilot decision rules for v270
Do not redesign from a single learner.
- Repeated friction from 3+ learners at the same interface boundary → product simplification candidate.
- Repeated conceptual misunderstanding after interaction → lesson / Enter-n-Action teaching candidate.
- Correct reasoning but difficulty showing evidence → evidence UX candidate.
- Smooth virtual diagnosis but weak physical transfer → practical-handoff / instructor-guidance candidate.
- No common pattern → gather more pilot sessions before changing architecture.

## Success signal
The pilot succeeds when learners can move from concept to measurement to diagnosis to evidence to physical handoff with minimal navigation coaching, and can explain **why** their test result supports or contradicts a diagnostic hypothesis.