What question did this study set out to answer?

The aim is to develop a sensor system that detects precursors to fires through thermal outgassing from cable insulation materials.

May 9, 2026Open Access

Development of a Pre-Ignition Fire Detection Sensor System

Key Points

The aim is to develop a sensor system that detects precursors to fires through thermal outgassing from cable insulation materials.
Investigated various materials (PMMA, Kapton, Nomex, PDMS) under ISS-like conditions using a thermogravimetric analyzer paired with a mass spectrometer.
Utilized thermally modulated gas sensor arrays based on semiconducting metal oxides to detect outgassing compounds.
Employed machine learning algorithms for binary classification to identify target gases against ambient air.
Detected specific thermal outgassing compounds from cable insulation at defined temperature thresholds.
Successfully distinguished gas composition patterns from ambient air and target gases using sensor arrays.
Optimized temperature modulation sequences significantly improved detection accuracy.

Abstract

Conventional smoke detectors on the International Space Station (ISS) work according to the scattered-light principle. Due to non-sedimenting dust in microgravity onboard, the detectors cause frequent false alarms, decreasing the crew's alertness for real danger and triggering unnecessary cleaning protocols. In other enclosed habitats such as a permanently occupied moon station this could even be made worse by an expected increase of moon dust over time. To prevent incipient fires instead of just detecting them, their sources need to be identified before ignition. As the primary origin is cable overheating, the goal is to recognize the presence of thermally outgassing compounds from the cable insulation. We report here a possible approach to detect these substances and consequently identify possible fire sources, using thermally modulated miniaturized gas sensor arrays -- based on semiconducting metal oxides. The sequence and shape of temperature modulation have proven to play a significant role in the detection of specific gases and can be optimized to the task at hand. In a previous feasibility study, power cable insulation materials were heated to different temperatures to distinguish between above and below maximum operating temperature. Currently we are adapting to ISS-like conditions: Different materials used on the ISS (PMMA, Kapton, Nomex, PDMS) are investigated using a thermogravimetric analyzer coupled with a mass spectrometer to confirm the alarm temperature threshold and gain information about the outgassing species. In a dedicated 1 m³ measurement setup, these materials are heated to temperatures close to the chosen thresholds and several sensor arrays measure the thermal exhaust. A machine learning algorithm is used to train with the gas composition-specific multidimensional output patterns to distinguish ambient air from the target gases caused by thermal outgassing. Using binary classification, the sensing performance is evaluated and the quality of the threshold detection quantified.

Bookmark

View Full Paper