Presence-Gated VOC Sensing for Urban Search and Rescue Applications

Read full text
Authors: Balasubramaniam, Kaveendran
Cited by: 0

Abstract

This dataset contains development-phase logs used to train and evaluate J48, Random Forest, and a simple probability-level hybrid classifier for a portable volatile-signature detector intended for urban search-and-rescue triage. Data were collected in tethered mode while the ESP32 firmware executed repeating intake–hold–purge cycles. During each hold window the device computed a per-hold feature vector and mirrored it to a host PC along with synchronized raw/derived channels and metadata.

Publication details

Journal / source
Mendeley Data
Publication date
9 February 2026
Document type
Dataset