Locating trapped or unconscious victims under rubble, behind walls or in smoky environments is a critical challenge
for first responders. While trained dogs (K9) excel at search
tasks, they are expensive and have limited availability. Recent
advances in commodity wireless hardware allow smartphones to
sense vital signs without line-of-sight by analyzing Wi-Fi channel
state information (CSI), Bluetooth Low Energy (BLE) received
signal strength (RSSI) and Ultra-Wideband (UWB) channel
impulse responses. Prior work on Wi-Fi sensing has shown
that scattering models can achieve robust performance even
with non-line-of-sight occlusion[1], and UWB radars can detect
human presence in NLOS conditions using machine learning[2].
We propose RF Biomarker Sense, a fused sensing pipeline that
combines CSI, BLE and UWB features on commodity phones to
detect human presence and estimate triage priority in occluded
scenarios. Our experiments demonstrate that fusion yields up to
a 12 % increase in area under the ROC curve (AUC) relative to
individual modalities and maintains operational detection under
occlusion distances up to 8 m. This paper describes the system
design, fusion algorithm and evaluation across distance, motion
and clutter conditions indoors and outdoors. Locating trapped or unconscious victims under rubble, behind walls or in smoky environments is a critical challenge
for first responders. While trained dogs (K9) excel at search
tasks, they are expensive and have limited availability. Recent
advances in commodity wireless hardware allow smartphones to
sense vital signs without line-of-sight by analyzing Wi-Fi channel
state information (CSI), Bluetooth Low Energy (BLE) received
signal strength (RSSI) and Ultra-Wideband (UWB) channel
impulse responses. Prior work on Wi-Fi sensing has shown
that scattering models can achieve robust performance even
with non-line-of-sight occlusion[1], and UWB radars can detect
human presence in NLOS conditions using machine learning[2].
We propose RF Biomarker Sense, a fused sensing pipeline that
combines CSI, BLE and UWB features on commodity phones to
detect human presence and estimate triage priority in occluded
scenarios. Our experiments demonstrate that fusion yields up to
a 12 % increase in area under the ROC curve (AUC) relative to
individual modalities and maintains operational detection under
occlusion distances up to 8 m. This paper describes the system
design, fusion algorithm and evaluation across distance, motion
and clutter conditions indoors and outdoors.