Speakers
Description
This paper presents a system-oriented experimental study on AI-enabled multi-sensor fusion for field detection of chemical warfare agents. The proposed framework integrates ion mobility spectrometry (IMS), portable Raman spectroscopy, and electrochemical sensing through feature-level fusion and interpretable machine learning. Its contribution is twofold: first, it formulates a practical fusion architecture for portable CBRN detection systems; second, it demonstrates a reproducible Python-based evaluation workflow on a realistic synthetic dataset derived from literature-reported response patterns for nerve-agent simulants, blister-agent simulants, choking-agent surrogates, and benign interferents. The fused model outperforms the individual sensing modalities, achieving 99.2% accuracy and a weighted F1-score of 0.992, compared with 94.7% for Raman-only, 92.8% for electrochemical-only, and 81.4% for IMS-only. The results indicate that combining complementary sensing physics with data-driven decision support improves discrimination robustness and reduces false alarms under environmental variability. The study is positioned as a reproducible intermediate step toward laboratory validation and future deployment-oriented CBRN sensing systems.