Elderly · Frontier
Taught-Home State Axes: Install-Day Personalization via Poke-and-Observe
In fifteen minutes on install day, the system learns 'fridge opened', 'shower ran', 'front door used'.
Plausible physics, early work — shown honestly as an experiment.
How it works
During setup, someone simply does each thing — opens the fridge, sits in the recliner, runs the shower — while the system watches how each action changes the WiFi field. Because it saw each action with a label, it can recognise them forever after. That turns vague 'activity' into 'Mom opened the fridge today'.
In the physics: Each controlled perturbation moves specific multipath components in a repeatable way: a metal fridge door swings a strong specular reflector through known angles (large CIR amplitude/delay signature); a front door changes the dominant hallway waveguide mode; shower water creates a transient high-eps_r/sigma volume. Because the perturbation is known and repeated, its signal subspace is directly identifiable against the ambient baseline.
See it work
This is real pipeline output for Elderly, replayed. The exact reading depends on your room.
Go deeper
How the field becomes an answer
How we prove it
Own-hardware experiment: 10 homes, 8 taught events each, measure per-axis detection F1 after 5 teaching repetitions and stability over 30 days (drift control over 119 days already proven supports baseline stability); no public dataset exists — this is a differentiating frontier capability.
Who it is for
For adult children buying peace of mind (D2C) and care coordinators: 'Teach the house what matters in 15 minutes — then get "she ate, she's up, the front door only opened for the nurse" every day, from WiFi alone.'
The matching solution
Complete first-room estimateAging-in-Place Guardian3 nodes · ~$130 parts · 60 min Request kit, subscribe, or set up →Compatible DIY hardware options
15 min · beginner
60 min · intermediate