Smorster

I Automated EVERY Minute of My Day

Smart Home Solver · 13m 00s · Watch on YouTube · 17 sources

Decision Card

Effort: A first automation from this video (motion-sensor lights, or opening shades before your phone alarm) is one weekend with a Home Assistant setup plus a $20–$80 sensor; replicating the whole house shown here is months of tinkering across dozens of devices.

Honest take: The video never names its platform or totals its cost — the “input Boolean” reference at [12:15] gives away that this is Home Assistant with heavy custom configuration, and the “self-healing” home is really one hand-built reboot automation per flaky device, not a general capability. Every automation is presented at demo-day polish with none of the setup, tuning, or false-trigger debugging shown.

Concrete next steps:

  • Pre-alarm wake-up (shade/light ramps before phone alarm) — try: uses the Android companion app’s Next Alarm sensor, ~1 evening; guide at Home Automation Guy. Skip on iOS — the sensor is Android-only.
  • Meeting mode (webcam-in-use locks door, sets DND, warns family) — adopt: camera_in_use binary sensor on macOS or HASS.Agent on Windows, 1–2 hours; worked example at jamesridgway.co.uk.
  • Self-healing device reboot (smart plug power-cycles offline devices) — adopt: cheapest win in the video, one automation per flaky device, under an hour each.
  • Room presence via Aqara FP2 mmWave sensor — try: ~$80 per room and zone calibration is fiddly, but it enables the still-body detection (dinner table, bed) that plain motion sensors can’t do.
  • CO2-triggered ventilation alerts — try: needs a $50–$150 CO2 sensor; the productivity claim has real research behind it (see Verified Claims).
  • Face-recognition door lock — skip unless already replacing a lock; onboard-face-unlock models like the Lockly Visage run ~$349.
  • Piano vibration sensor, sink pressure mat, plant-moisture lamp — skip: charming novelty tier, high maintenance-to-value ratio.

TL;DR

A smart home YouTuber walks through a full day where nearly every action — waking, showering, meetings, dinner, kids’ bedtime, lights-out — is handled by sensor-triggered automations (pressure mats, mmWave presence, vibration, CO2, webcam state) rather than voice commands or schedules. The recurring pattern worth stealing: use an existing signal you already generate (phone alarm, Bluetooth connect, camera-in-use, lying in bed) as the trigger, so the automation costs zero extra effort.

Key Points

  • Waking is a two-stage escalation: shades crack open minutes before the phone alarm, then a mattress pressure sensor detects lingering in bed and triggers full shades, an insult, and music 00:15
  • Connecting Bluetooth running headphones auto-starts a podcast playlist and texts his wife that he’s out running — the trigger is a device event, not a command 01:15
  • The front door lock scans his face on approach and unlocks hands-free 01:42
  • Flipping the shower light starts a warm-up timer shown as light-strip colors (green = ready, orange, red = get out) while a speaker reads the day’s calendar and to-do list 02:41
  • Meeting mode triggers off the computer camera being in use: door locks, hallway light turns red, phone goes to Do Not Disturb — and reverses when the camera turns off 03:44
  • The home “self-heals”: when a light strip drops offline, an automation power-cycles its smart outlet to bring it back 04:30
  • Package detection is inferred logic — doorbell sees a person AND the front door never opens — rather than a dedicated package sensor 05:00
  • High CO2 in the office flashes the light strips red and surfaces a dashboard button that turns on fans and opens the window shade 05:31
  • A school-pickup reminder fires when the garage door hasn’t opened in the last 10 minutes before pickup time — absence of an event as a trigger 06:09
  • An Aqara FP2 presence sensor detects when no one is still sitting at the dinner table and lifts Do Not Disturb automatically 08:27
  • A vibration sensor on the piano turns on lights when his daughter starts playing 09:03
  • A hallway motion sensor turns off the family-room TV when the kids get out of bed, so they don’t see what the parents are watching 10:30
  • Lying down in bed shuts down the whole house via the mattress pressure sensor — lights off, alarm armed, temporary input Booleans reset, plus a spoken warning if tomorrow’s first calendar event is earlier than the phone alarm 12:06

Notable Quotes

“will fully open up the shades give me a slight insult read get up and do something with your life” 00:46

“yeah my smart home self heals it’s amazing” 04:40

“it’s crazy how much more productive I am When there’s less CO2 in the room” 06:02

Verified Claims

  1. A mmWave presence sensor (Aqara FP2) can detect a person sitting still at a table and distinguish zones within a room. 08:27 — Aqara’s official specs confirm 60–64 GHz millimeter-wave radar, up to 30 customizable zones over ~40 m², and tracking of up to 5 people including stationary ones: Aqara Presence Sensor FP2, Amazon listing. Verdict: Confirmed.

  2. Elevated indoor CO2 measurably reduces productivity/cognitive performance. 06:02 — Harvard’s COGfx study (Allen et al., Environmental Health Perspectives) found ~15% lower cognitive scores at 950 ppm and ~50% at 1,400 ppm versus baseline: Yale Climate Connections summary, New Atlas on the follow-up office study. Effect sizes vary across studies, but the direction is well supported. Verdict: Confirmed.

  3. A smart lock can unlock automatically by scanning your face as you walk up. 01:44 — Commercially available; the Lockly Visage does onboard facial recognition with IR sensors, unlocking in ~1 second, with face data stored on-device: Lockly Visage product page, TechHive coverage. (The video doesn’t name its lock model.) Verdict: Confirmed.

  4. A smart home can know when your phone alarm is about to go off and act minutes beforehand. 00:20 — The Home Assistant Android companion app exposes a “Next Alarm” sensor; community blueprints trigger actions at a configurable offset before it. Not available on iOS: Home Automation Guy guide, HA community blueprint. Verdict: Confirmed.

  5. A smart home can detect that your computer’s camera is in use and trigger a meeting mode. 03:44 — Home Assistant’s macOS companion exposes a camera_in_use binary sensor, and HASS.Agent provides a Webcam Active sensor on Windows; both are commonly used to drive meeting lights: James Ridgway’s webcam-state writeup, How-To Geek on PC sensors in HA. Verdict: Confirmed.

  6. DeleteMe contacts data brokers to remove your info and sends periodic privacy reports with little ongoing effort from you. 07:21 — The sponsor’s mechanics check out (750+ broker sources, reports roughly every two months, ongoing re-monitoring), but independent reviews report only ~25–50% of records removed in the first months and note reappearing listings: DeleteMe privacy report docs, Cybernews review. Verdict: Confirmed with caveats — the service works as described but is not a complete scrub.

Tools, Papers & Standards Mentioned

Follow-up Questions

  1. What is the false-trigger rate of pressure/vibration/mmWave sensor automations in practice (e.g., pets, kids, laundry on the bed), and what debounce patterns do Home Assistant users apply to keep whole-house “bed shutdown” automations from misfiring?
  2. Can the pre-alarm wake-up pattern be replicated on iOS (where the Next Alarm sensor doesn’t exist) via Shortcuts-based alarm syncing, and how reliable is it?
  3. How well does the Harvard CO2-cognition effect replicate at typical home-office concentrations (800–1,200 ppm), and what CO2 sensor accuracy class (NDIR vs. eCO2) is needed for a ventilation automation to be meaningful?

Sources