Hospitality · Strong theory
TableState — restaurant table lifecycle tracking
Every table tells you its state: seated, eating, lingering, needs bussing.
The physics is solid; our own validation is scheduled.
How it works
Seated diners quietly block a little signal, and eating adds a gentle hand-to-mouth rhythm on top. When people leave, the moved chairs and dishes keep the table's signal fingerprint different from clean-and-empty. So the floor map knows which tables are done but not yet cleared — without a camera or a button.
In the physics: Seated bodies create static attenuation plus characteristic eating micro-Doppler (hand-to-mouth cycles ~0.2-0.5Hz arm motion, cutlery micro-motions); an abandoned but uncleared table shows the residual static multipath change of chairs/objects moved from baseline with no Doppler.
See it work
This is real pipeline output for Hospitality, replayed. The exact reading depends on your room.
Go deeper
How the field becomes an answer
How we prove it
OPERAnet activity classes as proxy for seated-activity discrimination; own 6-node ESP32 deployment over 8 tables in a partner restaurant with POS timestamps as ground truth.
Who it is for
Table-turn accelerator: tells the floor manager which tables are done but not bussed — sold on covers-per-night uplift to the restaurant GM.
runtime Runs today in LatentField 2.0 as machine_vibration
The matching solution
Complete first-room estimateFloor & Bar Pulse3 nodes · ~$130 parts · 90 min Request kit, subscribe, or set up →Compatible DIY hardware options
120 min · advanced