FLARE: An Explainable Habitat-Oriented Research Discovery and Decision-Support Framework for NASA Microgravity Combustion Investigations

Fire safety is a critical concern for crewed space exploration, and decades of microgravity combustion experiments have produced a large but fragmented body of evidence. A researcher who wants the evidence most relevant to a particular habitat scenario must still read many investigation records and compare atmospheres, fuels and metadata by hand. This paper presents FLARE (Fire Learning, Analysis & Risk Exploration), an early-stage, open-source Python prototype for scenario-driven discovery of NASA microgravity combustion investigations. A user defines a habitat scenario (oxygen concentration, total pressure, airflow, material category and fire-safety priorities). FLARE then ranks investigations with a transparent four-criterion weighted heuristic, reports a separate metadata-coverage value and a criterion-level breakdown for every result, supports side-by-side comparison, and reports catalogue-relative gaps. The prototype implements five interactive modules in Streamlit over an 11-record starter catalogue with an explicit JSON schema. We document the scoring model exactly as implemented and demonstrate it on reference scenarios. The demonstration exposes the central limitation of the current release: because no record yet contains a verified numerical oxygen, pressure or airflow value, rankings are identical for an ISS-like atmosphere (21% O2, 14.7 psia) and NASA’s exploration atmosphere (34% O2, 8.2 psia), every record has 45% coverage, and several records tie at a score of 100 when the user narrows the material and priority filters. We therefore present FLARE as a research-discovery workflow and design study, not a fire-risk model, and we specify the data curation and expert-relevance evaluation needed before its rankings can carry atmospheric meaning. Source code: https://github.com/naahi-rihan/FLARE

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-15
DOI
https://doi.org/10.5281/zenodo.24249938
Primary Topic
Fire dynamics and safety research
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

FLARE: An Explainable Habitat-Oriented Research Discovery and Decision-Support Framework for NASA Microgravity Combustion Investigations

Nahi Mumtaj
Zenodo (CERN European Organization for Nuclear Research)
Fire dynamics and safety research
preprint

FLARE: An Explainable Habitat-Oriented Research Discovery and Decision-Support Framework for NASA Microgravity Combustion Investigations

Nahi Mumtaj
preprint en

Abstract

Fire safety is a critical concern for crewed space exploration, and decades of microgravity combustion experiments have produced a large but fragmented body of evidence. A researcher who wants the evidence most relevant to a particular habitat scenario must still read many investigation records and compare atmospheres, fuels and metadata by hand. This paper presents FLARE (Fire Learning, Analysis & Risk Exploration), an early-stage, open-source Python prototype for scenario-driven discovery of NASA microgravity combustion investigations. A user defines a habitat scenario (oxygen concentration, total pressure, airflow, material category and fire-safety priorities). FLARE then ranks investigations with a transparent four-criterion weighted heuristic, reports a separate metadata-coverage value and a criterion-level breakdown for every result, supports side-by-side comparison, and reports catalogue-relative gaps. The prototype implements five interactive modules in Streamlit over an 11-record starter catalogue with an explicit JSON schema. We document the scoring model exactly as implemented and demonstrate it on reference scenarios. The demonstration exposes the central limitation of the current release: because no record yet contains a verified numerical oxygen, pressure or airflow value, rankings are identical for an ISS-like atmosphere (21% O2, 14.7 psia) and NASA’s exploration atmosphere (34% O2, 8.2 psia), every record has 45% coverage, and several records tie at a score of 100 when the user narrows the material and priority filters. We therefore present FLARE as a research-discovery workflow and design study, not a fire-risk model, and we specify the data curation and expert-relevance evaluation needed before its rankings can carry atmospheric meaning. Source code: https://github.com/naahi-rihan/FLARE

Zenodo (CERN European Organization for Nuclear Research)
Fire dynamics and safety research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon