Exploring pre-ignition source localization using a 3D network of semiconducting metal oxide gas sensors

Early fire detection is critical, particularly given the increasing use of plastics that release toxic gases during overheating. Despite its importance, there is limited literature on combining pre-ignition fire detection with spatial sensor networks capable of localizing the source. In this work, we investigated the ability of a network of 32 commercial semiconducting metal oxide sensors in a 1 m³ chamber to differentiate between three possible locations of heated PMMA as a proof-of-concept. The study also provides insights into plume dynamics and highlights the importance of sensor placement for accurate source differentiation. The results demonstrate the potential of semiconducting metal oxide gas sensor networks for early-warning fire detection in indoor environments.