CLASSROOM // HOMELAB ACADEMY M4 LAB
M4 LABlab

Homelab Academy / Modules

Hands-on 01-voice-architecture: Lab

Guided practice + break/fix + lab gate.

Show commands for:
Class 11: Local voice architecture
Class 11 visual reference. Use it as the map, then prove it in the reading, lesson, and lab.

Overview

Module: Module 4: Local Voice Assistant
Activity type: Lab (Practice)
Objective: Given a voice request, the learner can draw the local Assist pipeline stages, instrument one test per stage, and isolate faults to mic, wake, STT, intent/action, TTS, or playback.

Learn

Before you start

  • Reading complete
  • Lesson Feynman teach-back drafted (you may refine after the lab)
Hands-on

Guided lab

Test the voice chain one layer at a time. Do not run full wake→action until layers 1–10 pass.

1

Record raw mic audio (arecord/Audacity/Phone). Confirm intelligible, not clipped.

Linux / macOS (Terminal)
# Example if ALSA device exists on the satellite/host:
arecord -l
arecord -d 5 -f cd /tmp/voice-raw.wav
aplay /tmp/voice-raw.wav
ls -l /tmp/voice-raw.wav
Windows (CMD / PowerShell)
# Record with Voice Recorder / Audacity; save to Desktop\voice-raw.wav
Get-Item $HOME\Desktop\voice-raw.wav | Format-List Name, Length, LastWriteTime
2

Wake word: 10 attempts at fixed distance → record detections / misses.

3

False activations: 10 unrelated phrases → record false wakes.

4

VAD/endpointing: confirm start/end do not chop words (inspect recording or pipeline debug).

5

STT alone: submit the same audio/phrase to configured STT; save transcript text in workbook.

6

Assist text path: paste that exact transcript into HA Assist; record intent + target entity.

7

Action alone: call the HA service manually; confirm entity changes.

8

TTS alone: synthesize a fixed phrase; confirm audio file/stream exists.

9

Playback alone: play a known WAV/MP3 on the target speaker.

Linux / macOS (Terminal)
aplay /usr/share/sounds/alsa/Front_Center.wav 2>/dev/null || ffplay -autoexit -nodisp /tmp/test.wav
10

Full pipeline only after 1–9 pass. Time wake→action latency; log failures by layer.

Break / fix

Break / fix

Break/fix

Mute mic; confirm wake fails. Unmute. Play TTS with speaker unplugged; confirm synthesis may succeed while playback fails, record both.

Fix chart

Feedback / common mistakes

  • Skipping evidence and calling it done.
  • Changing multiple variables before retesting.
  • Moving on without completing Feynman teach-back.
Hands-on

Lab gate

All boxes must be true before the next class unlocks. If you fail: feedback → targeted review → new practice → reassess.

  • [ ] Raw microphone sample is acceptable.
  • [ ] Wake-word hit/miss evidence is recorded.
  • [ ] STT transcript meets criterion.
  • [ ] Text-only HA command works.
  • [ ] TTS and speaker pass separately.
  • [ ] Stage timestamps identify latency.
  • [ ] End-to-end command passes.
  • [ ] Prior-knowledge check answered
  • [ ] Feynman teach-back completed (Explain, Simplify, Example, Weak spot, Retry)
  • [ ] Independent practice completed
  • [ ] Retrieval check attempted (target ≥80% when scored)
  • [ ] Evidence recorded in workbook / verification matrix
← LessonHomework →