Problem with the Other Sensors

Passive-infrared (PIR) sensors miss still occupants → WaveSense detects
chest-rise micro-motions (< 0.25 mm).• Vision analytics trigger privacy concerns → WaveSense is 100 % anonymous RF
(captures no images or PII).• Bluetooth beacons rely on visitor badges → WaveSense is device-free and
badge-free.• Multi-sensor installs are complex → One sensor covers up to 10 m (33 ft)
radius.

How WaveSense Works

60 GHz mmWave front-end
– Gigahertz-wide bandwidth resolves sub-millimetre movements.On-sensor DSP
– Real-time FFT and clutter suppression shrink raw data on the edge.Neural-network inference
– TinyML models classify presence, count heads and can detect breathing.Edge / Cloud API
– JSON via REST, MQTT or BACnet. Latency < 50 ms on-prem; < 300 ms cloud.

Key Capabilities

Still-presence detection : Motion ≥ 0.25 mm → {"presence": true}
Occupancy counting : 1–10 people @ 10 m → {"count": 4}
Vital signs (beta) : ± 1 breath/min → {"respiration": 18}
Through-wall sensing : 15 cm drywall → {"presence": true}98.3 % still-presence accuracy in office pilot
• 0.6 W typical power draw at 4 Hz updates
• 20 dB signal-to-noise advantage over Wi-Fi CSI alone( Two simple line charts would sit here: Accuracy vs Distance and
Power vs Update Rate )