Abstract:With air quality being one target in the sustainable development goals set by the United Nations, accurate monitoring also of indoor air quality is more important than ever. Chemiresistive gas sensors are an inexpensive and promising solution for the monitoring of volatile organic compounds, which are of high concern indoors. To fully exploit the potential of these sensors, advanced operating modes, calibration, and data evaluation methods are required. This contribution outlines a systematic approach based on dynamic operation (temperature-cycled operation), randomized calibration (Latin hypercube sampling), and the use of advances in deep neural networks originally developed for natural language processing and computer vision, applying this approach to volatile organic compound measurements for indoor air quality monitoring. This paper discusses the pros and cons of deep neural networks for volatile organic compound monitoring in a laboratory environment by comparing the quantification accuracy of state-of-the-art data evaluation methods with a 10-layer deep convolutional neural network (TCOCNN). The overall performance of both methods was compared for complex gas mixtures with several volatile organic compounds, as well as interfering gases and changing ambient humidity in a comprehensive lab evaluation. Furthermore, both were tested under realistic conditions in the field with additional release tests of volatile organic compounds. The results obtained during field testing were compared with analytical measurements, namely the gold standard gas chromatography mass spectrometry analysis based on Tenax sampling, as well as two mobile systems, a gas chromatograph with photo-ionization detection for volatile organic compound monitoring and a gas chromatograph with a reducing compound photometer for the monitoring of hydrogen. The results showed that the TCOCNN outperforms state-of-the-art data evaluation methods, for example for critical pollutants such as formaldehyde, achieving an uncertainty of around 11 ppb even in complex mixtures, and offers a more robust volatile organic compound quantification in a laboratory environment, as well as in real ambient air for most targets.

Label-free Bio-Aerosol Detection and Classification Using a Virtual Impactor and Deep Learning (conference Presentation)

Virtual impactor-based label-free bio-aerosol detection using holography and deep learning

Label-Free Bioaerosol Sensing Using Mobile Microscopy and Deep Learning

System Integration of an Optimally Designed Virtual Impactor with a QCM Sensor As a One-Stop PM2.5 Classification and Detection Platform

Label-free Bio-aerosol Sensing Using On-Chip Holographic Microscopy and Deep Learning

Pollen Detection Through Integrated Microfluidics and Smartphone-Driven Deep Learning Systems

Real Time and Online Aerosol Identification Based on Deep Learning of Multi-Angle Synchronous Polarization Scattering Indexes.

A virtual impactor-based portable optical sensor with probabilistic neural network: for particle property identification

Atmospheric aerosol diagnostics with UAV-based holographic imaging and computer vision

An Integrated Air Quality Monitoring System for COVID-19 Virus Detection

Real-time detection of viruses in aerosols with weak light imaging based on Tesla discharge

In situ Biological Particle Analyzer based on Digital Inline Holography

Online Detection of Airborne Nanoparticle Composition with Mass Spectrometry: Recent Advances, Challenges, and Opportunities

Recent Advances on Bioaerosol Collection and Detection in Microfluidic Chips

Dual-modal Measurements of Suspended Particles Combining Polarization and Fluorescence Analysis

High-Performance VOC Quantification for IAQ Monitoring Using Advanced Sensor Systems and Deep Learning

Detection of airborne pathogens with single photon counting and a real-time spectrometer on microfluidics

Single aerosol recognition based on deep learning of multidimensional polarization signals

Deep learning enables high-throughput analysis of particle-aggregation-based bio-sensors imaged using holography

Real-time detection of molecular structures and metal components within airborne particles using concurrent Raman and Spark emission spectroscopy

Deep-AIR: A Hybrid CNN-LSTM Framework for Air Quality Modeling in Metropolitan Cities