The dirty secret of home automation is that the person who sets it up is usually the only person who touches it. You can wire your house with the cleverest automations on earth — weather-aware sprinklers, presence-based HVAC, blinking-light timers — and your household will still walk past the tablet on the wall like it's a fire extinguisher. The author of this project understood that the problem was never technical. It was aesthetic. The project's core move is simple and a little inspired: commission a professional environment artist to draw your house, then use those illustrations as the literal interface layer for Home Assistant's picture-elements card. Every device gets hand-drawn on and off states. Day and night versions swap at sunset. Animations are eight-frame WebP loops — the fireplace flickers, the AC breathes. The technical stack underneath is stock Home Assistant with a few HACS add-ons (Browser Mod for custom dialogs, card-mod for CSS positioning, HAOS Kiosk Display for TV output). Nothing exotic. The illustration IS the interface. What makes this more than a pretty screenshot is the production pipeline the author documents. The brief to the artist is a small masterpiece of technical communication: architectural plans with furniture marked in, room photos, style references, paint codes from the last renovation, and — critically — the instruction that every asset must be exported on a shared canvas at the same coordinates. That single constraint eliminates days of pixel-alignment hell. The author learned this the hard way and is saving you the trip. The YAML architecture is quietly clever. Each floor's elements live in a separate include file, referenced by both the individual floor view and the combined "Full" view. Edit a room once, it updates everywhere. The full-house view reconstructs the shared canvas — floors snap back to their original positions like layers in Photoshop, with a slight vertical offset to create an exploded-diagram effect. It's the kind of solution that looks obvious after someone else figures it out. The TV deployment story is a useful reality check. The author wanted the dashboard always visible on the living room TV. webOS wouldn't cooperate — no custom screensaver, no default browser page. The solution was to bypass the TV's software entirely: plug the Home Assistant box into HDMI via the Kiosk Display add-on. Physical problems get physical solutions. The aside about WebP support on LG TVs is a gift to anyone who's been scared off by outdated forum posts — it works fine in the browser, the limitation was only in the photo viewer app. The project's real contribution isn't the code or the cards — it's the workflow. Finding a specialist environment artist, writing a brief that bridges the gap between "commission" and "software interface asset," structuring deliverables so the technical integration isn't miserable. The author makes a convincing case that the hardest part of the project was everything before the first line of YAML. That's an honest and useful observation for anyone attempting something similar. This is a blog post, not a product, and it's worth evaluating as a piece of craft writing about a craft project. The writing is clean, the technical detail is specific without being exhausting, and the project itself solves a real problem — household adoption — through a genuinely creative approach. It's the kind of work that makes you want to build something.
