Rapid and reliable discrimination of authentic nerve agents in complex environments is a critical, unsolved challenge for security and public safety. Current detection technologies are often limited by bulky instrumentation, prolonged analysis times, or validation restricted to simulants, hindering their real-world threat-monitoring capabilities. Here, we demonstrate a machine learning-enhanced colorimetric sensor array that overcomes these limitations. We employed a systematic data-driven approach using hierarchical clustering, analysis of variance, and correlation analysis to optimally select just six commercially available fluorescent dyes from an initial 29-candidate library. This optimized array was rigorously evaluated against five authentic nerve agents (GA, GB, GD, GF, VX) and a key simulant using a novel dual-mode (Visible/UV) illumination strategy. By quantifying colorimetric responses (RGB-to-ΔE₀₀), our linear discriminant analysis classifier achieved 100% classification accuracy. The sensor provides an instantaneous visual response and maintains discrimination capability down to 10 μM, markedly outperforming the 87.5% accuracy of visible-light-only detection. This study establishes a simple, low-cost, and scalable platform for the rapid, accurate discrimination of multiple authentic nerve agents. By integrating data-driven sensor design with robust visual analysis, our system provides a clear pathway toward next-generation portable technologies for real-time chemical threat surveillance.

The rapid detection and selective discrimination of chemical warfare agents (CWAs) remain a critical unmet challenge for field–deployable sensing technologies. Each CWA class (blood, choking, blister, etc.) shares common structural motifs, yet specific identification of individual members is essential for real–time hazard assessment. Conventional techniques such as gas chromatography-mass spectrometry (GC–MS), ion mobility spectrometry (IMS), and Fourier-transform infrared spectroscopy (FT-IR) are precise but rely on bulky instrumentation and trained operators, limiting their use in emergency scenarios. Herein, we report a machine learning–assisted colorimetric/fluorometric sensor array for classifying seven representative CWAs (hydrogen cyanide, cyanogen chloride, phosgene, chloropicrin, sulfur mustard, nitrogen mustards, lewisites). The platform employs 29 commercially available dyes that generate distinct optical fingerprints under visible and UV illumination (λ = 365 nm), captured by a smartphone–based imaging system and quantified as CIE Lab coordinates. Systematic statistical optimization (HCA, ANOVA, correlation analysis) reduced redundancy and identified an informative subset, while multivariate methods (LDA, t–SNE) confirmed robust separation across concentration ranges. The optimized array exhibited high accuracy, reproducibility, and portability, underscoring its potential as a cost–effective tool for on–site CWA monitoring.

Resource-constrained edge processors deployed on unmanned aerial vehicles and wearable platforms require compact, drift-robust gas classification models for a range of environmental and security monitoring applications, including CBRN-motivated scenarios. Existing approaches rely on server-grade architectures incompatible with edge-board-scale deployment, or on classifiers that chemically degrade severely under long-term sensor drift. Each UCI gas class was mapped to a CBRN behavioral category based on physicochemical analogy (molecular functional group, vapor pressure, and metal-oxide semiconductor (MOS) cross-sensitivity pattern), following established precedent. Analyzed were Ammonia (NH₃), Acetaldehyde (CH₃CHO), Acetone ((CH₃)₂CO), Ethylene (C₂H₄), Ethanol (C₂H₅OH), Toluene (C₆H₅CH₃). We propose herein an end-to-end pipeline integrating a novel 1-D convolutional neural network with depth-wise separable convolutions (LiteSensor-Net), INT8 post-training quantization, structured magnitude pruning, and a knowledge-distillation domain-adaptation module (KD–DM) for sensor drift compensation. Using the UCI Gas Sensor Array Drift Dataset (13,910 measurements; 16 metal-oxide sensors; six analyte gases; a 36-month work span). LiteSensor-Net achieved accuracy = 92.63 ± 2.02%, macro-F1 = 0.898, model size = 5.99 kB INT8 pruned, inference latency = 6.3 ms, RAM footprint = 31.7 kB, and energy per inference = 0.04 mJ (all metrics on Raspberry Pi 4B, ARM Cortex-A72). Under chronological forward-chaining evaluation, KD–DM–20 achieved 47.91 ± 18.79% mean accuracy over Batches 2–10, representing a +9.25 pp improvement over uncompensated NC (38.66%). A six-metric benchmark framework—accuracy, macro-F1, model size, inference latency, RAM footprint, and energy per inference—is introduced to standardize edge-AI gas classifier evaluation. The proposed pipeline provides an open-source, deployable foundation for edge-class gas classification systems, with CBRN detection as a motivating application. Full operational validation on certified chemical simulants remains as future work.

The rapid expansion of the industrial chemical space has outpaced regulatory oversight, creating a critical regulatory and environmental blind spot regarding unregulated high-consequence chemicals with limited hazard characterization. While conventional QSAR models focus on predictive accuracy for known chemicals, they often lack the decision-oriented screening required for environmental risk management. Here, we present a "proactive discovery" framework designed to unveil hidden high-consequence threats among 16,000+ structural analogs of toxic industrial chemicals (TICs). The central contribution of this work is the integration of multiple hazard endpoints, including acute, dermal, ocular, respiratory, flammability, reactivity, and aquatic toxicity, onto a single regulatory-aligned 0-4 scale, rather than the prediction of any individual endpoint. By integrating National Fire Protection Association (NFPA) and Globally Harmonized System (GHS) standards into a unified hazard scale, we enable consistent risk prioritization across multiple endpoints. Strictly validated against an independent external set of 112 compounds, the framework demonstrated a "safety-first" predictive capability. To address environmental safety blind spots, the model was optimized to prioritize sensitivity for the highest-hazard category (Class 4). This strategy minimizes false negatives, ensuring that potential ecotoxicological hazards and high-consequence threats are not overlooked in environmental risk assessments. Crucially, this study moves beyond method development to actionable discovery. Applied to more than 16,000 structural analogs of toxic industrial chemicals, this multi-endpoint prioritization identifies a small, actionable subset of unregulated compounds that are classified at the maximum hazard level in two or more endpoints simultaneously, and are therefore strong candidates for immediate regulatory attention. We identified 20 high-priority unregulated candidates, including novel isocyanates and organolead analogs, that exhibit "Dual-Threats" of extreme acute toxicity to mammals and severe risks to aquatic ecosystems. These candidates display hazard profiles that warrant comparison to highly regulated toxic agents (including chemical warfare agents) for precautionary prioritization. Furthermore, our feature importance analysis reveals key molecular determinants driving toxicity, providing mechanistic support for these high-hazard predictions. By openly providing a prioritized list of 20 unregulated high-consequence agents, this study offers an actionable roadmap for policymakers and experimentalists to preemptively secure the chemical safety landscape against emerging high-consequence chemicals.