Mapping Occupational Hazards with a Multi-sensor Network in a Heavy-Vehicle Manufacturing Facility

Design and testing of a personalized noise monitoring system

Agricultural workers are more prone to noise-induced hearing loss than are many other workers. Hearing protection device use among agricultural workers is low, but training can increase hearing protection device use. This work proposes a system designed to automatically inform agricultural workers when they were exposed to noises that exceed the National Institute for Occupational Safety and Health (NIOSH) recommended exposure level. The smartphone-based system worn on the arm uses a noise dosimeter to measure noise exposures throughout the day to within ±2 A-weighted decibels of a Class 2 sound level meter. The device collects location and audio data, which are transferred to a server and presented to the worker on a locally hosted website. The website details noise exposure and helps the worker identify where exposure occurred and what specific tasks exceed NIOSH's recommended exposure limit, putting them at higher risk of noise-induced hearing loss. With this understanding, the worker is expected to adopt behavior changes and better hearing protection use at critical places and times. This pilot study evaluates the accuracy of the noise dosimeter and GPS relative to gold-standard instruments. The system was tested on a farm with outputs compared with gold-standard instruments. A-weighted, 1-sec averaged sound pressure levels and position data were collected while users were performing a variety of tasks indoors and outdoors. The smartphone's external noise dosimeter read within ±2 A-weighted decibels of the Class 2 reference dosimeter 59% of the time. The positioning devices had an average error of sub-4 m. While not perfectly matching gold-standard instruments, the device is capable of identifying and collecting information relative to loud noise events that promote noise-induced hearing loss.

Intra-workday fluctuations of airborne contaminant concentration and the time-weighted average

Air contaminant concentrations vary between and within workdays and are often measured across a workday by passing a known air volume through a collection device. Laboratory analysis determines the contaminant mass trapped, providing a time-weighted average air concentration (C_TWA). This approach was driven by the best technologies available as exposure measurement processes developed and accuracy and measurement precision were sought. However, all integrated concentration•time (C•t) values determining C_TWA are equally weighted in assessing exposures, intra-workday concentration variability is unknown, and results are available days later. At times inappropriately, an occupational exposure limit (OEL) expressed as a C_TWA also requires equal weighting of all C•t values across an exposure period following concepts of Haber's law. Continuous monitoring (real-time detection) informs both the C_TWA and the variability of C during sampling, which are needed for stressors where a ceiling or peak OEL exists, for dangerous exposures to permanent gas-type contaminants, and for immediately dangerous to life or health (IDLH) conditions. Selective and accurate real-time detection instruments are not available for all air contaminants, but exposure magnitude information may be provided. The large amounts of data from continuous monitoring and the ability to correlate exposure maxima to specific tasks are also important. An exposure assessment role exists for selective and nonselective monitors, and in some cases, similar accuracy and precision are provided compared to laboratory analyses. Continuous monitoring may be of value when the alternative is the collection of a few C_TWA data points. Digitized personal monitor data can support the automation of some exposure control decisions or allow such decisions to be made by people in near real-time. The emerging Internet of Things (IoT) offers opportunities to integrate digital exposure data into decision-making to increase both efficiency and safety. The perceived and real uncertainty associated with real-time exposure assessments may be lessened with work to rule out the presence of know interferents and confirm the presence of target analytes.

Direct-reading instruments for aerosols: A review for occupational health and safety professionals part 2: Applications

Direct reading instruments (DRIs) for aerosols have been used in industrial hygiene practice for many years, but their potential has not been fully realized by many occupational health and safety professionals. Although some DRIs quantify other metrics, this article will primarily focus on DRIs that measure aerosol number, size, or mass. This review addresses three applications of aerosol DRIs that occupational health and safety professionals can use to discern, characterize, and document exposure conditions and resolve aerosol-related problems in the workplace. The most common application of aerosol DRIs is the evaluation of engineering controls. Examples are provided for many types of workplaces and situations including construction, agriculture, mining, conventional manufacturing, advanced manufacturing (nanoparticle technology and additive manufacturing), and non-industrial sites. Aerosol DRIs can help identify the effectiveness of existing controls and, as needed, develop new strategies to reduce potential aerosol exposures. Aerosol concentration mapping (ACM) using DRI data can focus attention on emission sources in the workplace spatially illustrate the effectiveness of controls and constructively convey concerns to management and workers. Examples and good practices of ACM are included. Video Exposure Monitoring (VEM) is another useful technique in which video photography is synced with the concentration output of an aerosol DRI. This combination allows the occupational health and safety professional to see what tasks, environmental situations, and/or worker actions contribute to aerosol concentration and potential exposure. VEM can help identify factors responsible for temporal variations in concentration. VEM can assist with training, engage workers, convince managers about necessary remedial actions, and provide for continuous improvement of the workplace environment. Although using DRIs for control evaluation, ACM and VEM can be time-consuming, the resulting information can provide useful data to prompt needed action by employers and employees. Other barriers to adoption include privacy and security issues in some worksites. This review seeks to provide information so occupational health and safety professionals can better understand and effectively use these powerful applications of aerosol DRIs.

Direct-reading instruments for aerosols: A review for occupational health and safety professionals part 1: Instruments and good practices

With advances in technology, there are an increasing number of direct-reading instruments available to occupational health and safety professionals to evaluate occupational aerosol exposures. Despite the wide array of direct-reading instruments available to professionals, the adoption of direct-reading technology to monitor workplace exposures has been limited, partly due to a lack of knowledge on how the instruments operate, how to select an appropriate instrument, and challenges in data analysis techniques. This paper presents a review of direct-reading aerosol instruments available to occupational health and safety professionals, describes the principles of operation, guides instrument selection based on the workplace and exposure, and discusses data analysis techniques to overcome these barriers to adoption. This paper does not cover all direct-reading instruments for aerosols but only those that an occupational health and safety professional could use in a workplace to evaluate exposures. Therefore, this paper focuses on instruments that have the most potential for workplace use due to their robustness, past workplace use, and price with regard to return on investment. The instruments covered in this paper include those that measure aerosol number concentration, mass concentration, and aerosol size distributions.

Evolution and Applications of Recent Sensing Technology for Occupational Risk Assessment: A Rapid Review of the Literature

Over the last decade, technological advancements have been made available and applied in a wide range of applications in several work fields, ranging from personal to industrial enforcements. One of the emerging issues concerns occupational safety and health in the Fourth Industrial Revolution and, in more detail, it deals with how industrial hygienists could improve the risk-assessment process. A possible way to achieve these aims is the adoption of new exposure-monitoring tools. In this study, a systematic review of the up-to-date scientific literature has been performed to identify and discuss the most-used sensors that could be useful for occupational risk assessment, with the intent of highlighting their pros and cons. A total of 40 papers have been included in this manuscript. The results show that sensors able to investigate airborne pollutants (i.e., gaseous pollutants and particulate matter), environmental conditions, physical agents, and workers' postures could be usefully adopted in the risk-assessment process, since they could report significant data without significantly interfering with the job activities of the investigated subjects. To date, there are only few "next-generation" monitors and sensors (NGMSs) that could be effectively used on the workplace to preserve human health. Due to this fact, the development and the validation of new NGMSs will be crucial in the upcoming years, to adopt these technologies in occupational-risk assessment.

Evaluating the Performance of Using Low-Cost Sensors to Calibrate for Cross-Sensitivities in a Multipollutant Network

As part of our low-cost sensor network, we colocated multipollutant monitors containing sensors for particulate matter, carbon monoxide, ozone, nitrogen dioxide, and nitrogen monoxide at a reference field site in Baltimore, MD, for 1 year. The first 6 months were used for training multiple regression models, and the second 6 months were used to evaluate the models. The models produced accurate hourly concentrations for all sensors except ozone, which likely requires nonlinear methods to capture peak summer concentrations. The models for all five pollutants produced high Pearson correlation coefficients (r > 0.85), and the hourly averaged calibrated sensor and reference concentrations from the evaluation period were within 3-12%. Each sensor required a distinct set of predictors to achieve the lowest possible root-mean-square error (RMSE). All five sensors responded to environmental factors, and three sensors exhibited cross-sensitives to another air pollutant. We compared the RMSE from models (NO₂, O₃, and NO) that used colocated regulatory instruments and colocated sensors as predictors to address the cross-sensitivities to another gas, and the corresponding model RMSEs for the three gas models were all within 0.5 ppb. This indicates that low-cost sensor networks can yield useable data if the monitoring package is designed to comeasure key predictors. This is key for the utilization of low-cost sensors by diverse audiences since this does not require continual access to regulatory grade instruments.

Probabilistic Machine Learning with Low-Cost Sensor Networks for Occupational Exposure Assessment and Industrial Hygiene Decision Making

Occupational exposure assessments are dominated by small sample sizes and low spatial and temporal resolution with a focus on conducting Occupational Safety and Health Administration regulatory compliance sampling. However, this style of exposure assessment is likely to underestimate true exposures and their variability in sampled areas, and entirely fail to characterize exposures in unsampled areas. The American Industrial Hygiene Association (AIHA) has developed a more realistic system of exposure ratings based on estimating the 95th percentiles of the exposures that can be used to better represent exposure uncertainty and exposure variability for decision-making; however, the ratings can still fail to capture realistic exposure with small sample sizes. Therefore, low-cost sensor networks consisting of numerous lower-quality sensors have been used to measure occupational exposures at a high spatiotemporal scale. However, the sensors must be calibrated in the laboratory or field to a reference standard. Using data from carbon monoxide (CO) sensors deployed in a heavy equipment manufacturing facility for eight months from August 2017 to March 2018, we demonstrate that machine learning with probabilistic gradient boosted decision trees (GBDT) can model raw sensor readings to reference data highly accurately, entirely removing the need for laboratory calibration. Further, we indicate how the machine learning models can produce probabilistic hazard maps of the manufacturing floor, creating a visual tool for assessing facility-wide exposures. Additionally, the ability to have a fully modeled prediction distribution for each measurement enables the use of the AIHA exposure ratings, which provide an enhanced industrial decision-making framework as opposed to simply determining if a small number of measurements were above or below a pertinent occupational exposure limit. Lastly, we show how a probabilistic modeling exposure assessment with high spatiotemporal resolution data can prevent exposure misclassifications associated with traditional models that rely exclusively on mean or point predictions.

A Scoping Review of Technologies and Their Applicability for Exposome-Based Risk Assessment in the Oil and Gas Industry

INTRODUCTION

Oil and gas workers have been shown to be at increased risk of chronic diseases including cancer, asthma, chronic obstructive pulmonary disease, and hearing loss, among others. Technological advances may be used to assess the external (e.g. personal sensors, smartphone apps and online platforms, exposure models) and internal exposome (e.g. physiologically based kinetic modeling (PBK), biomonitoring, omics), offering numerous possibilities for chronic disease prevention strategies and risk management measures. The objective of this study was to review the literature on these technologies, by focusing on: (i) evaluating their applicability for exposome research in the oil and gas industry, and (ii) identifying key challenges that may hamper the successful application of such technologies in the oil and gas industry.

METHOD

A scoping review was conducted by identifying peer-reviewed literature with searches in MEDLINE/PubMed and SciVerse Scopus. Two assessors trained on the search strategy screened retrieved articles on title and abstract. The inclusion criteria used for this review were: application of the aforementioned technologies at a workplace in the oil and gas industry or, application of these technologies for an exposure relevant to the oil and gas industry but in another occupational sector, English language and publication period 2005-end of 2019.

RESULTS

In total, 72 articles were included in this scoping review with most articles focused on omics and bioinformatics (N = 22), followed by biomonitoring and biomarkers (N = 20), external exposure modeling (N = 11), PBK modeling (N = 10), and personal sensors (N = 9). Several studies were identified in the oil and gas industry on the application of PBK models and biomarkers, mainly focusing on workers exposed to benzene. The application of personal sensors, new types of exposure models, and omics technology are still in their infancy with respect to the oil and gas industry. Nevertheless, applications of these technologies in other occupational sectors showed the potential for application in this sector.

DISCUSSION AND CONCLUSION

New exposome technologies offer great promise for personal monitoring of workers in the oil and gas industry, but more applied research is needed in collaboration with the industry. Current challenges hindering a successful application of such technologies include (i) the technological readiness of sensors, (ii) the availability of data, (iii) the absence of standardized and validated methods, and (iv) the need for new study designs to study the development of disease during working life.

Real-time air monitoring of occupational exposures to particulate matter among hairdressers in Maryland: A pilot study

Hairdressers are exposed to particulate matter (PM), a known air pollutant linked to adverse health effects. Still, studies on occupational PM exposures in hair salons are sparse. We characterized indoor air PM concentrations in three salons primarily serving an African/African American (AA) clientele, and three Dominican salons primarily serving a Latino clientele. We also assessed the performance of low-cost sensors (uRAD, Flow, AirVisual) by comparing them to high-end sensors (DustTrak) to conduct air monitoring in each salon over 3 days to quantify work shift concentrations of PM_2.5 , respirable PM (RPM), and PM₁₀ . We observed high spatial and temporal variability in 30-min time-weighted average (TWA) RPM concentrations (0.18-5518 μg/m³ ). Readings for the uRAD and AirVisual sensors were highly correlated with the DustTrak (R²  = 0.90-0.99). RPM 8-hour TWAs ranged from 18 to 383 µg/m³ for AA salons, and 9-2115 µg/m³ for Dominican salons. Upper 95th percentiles of daily RPM exposures ranged from 439 to 2669 µg/m³ . The overall range of 30-min TWA PM_2.5 and PM₁₀ concentrations was 0.13-5497 and 0.36-,541 μg/m³ , respectively. Findings suggest that hairdressers could be overexposed to RPM during an 8-hour shift. Additional comprehensive monitoring studies are warranted to further characterize temporal and spatial variability of PM exposures in this understudied occupational population.

Predicting nurse fatigue from measures of work demands

Fatigue arising from excessive work demands is a known safety challenge in hospital nurses. This study aimed to determine which measures of work demands during nursing work are most predictive of hospital nurse fatigue levels at the end of the work shift. Measures of work demands of registered nurses from two hospital units in the United States were collected from organizational data sources, wearable sensors, and questionnaires. Fatigue levels were measured at the start and end of each shift using the Brief Fatigue Inventory. Multilevel linear regression analysis was used to predict end of shift fatigue based on work demand variables. The best fit model included multiple variables from organizational data sources and a physical activity variable measured by a wearable sensor. Organizational data can be used to create dynamic measures of work demands as they occur and predict end of shift fatigue levels in hospital nurses.

Estimating personal exposures from a multi-hazard sensor network

Occupational exposure assessment is almost exclusively accomplished with personal sampling. However, personal sampling can be burdensome and suffers from low sample sizes, resulting in inadequately characterized workplace exposures. Sensor networks offer the opportunity to measure occupational hazards with a high degree of spatiotemporal resolution. Here, we demonstrate an approach to estimate personal exposure to respirable particulate matter (PM), carbon monoxide (CO), ozone (O₃), and noise using hazard data from a sensor network. We simulated stationary and mobile employees that work at the study site, a heavy-vehicle manufacturing facility. Network-derived exposure estimates compared favorably to measurements taken with a suite of personal direct-reading instruments (DRIs) deployed to mimic personal sampling but varied by hazard and type of employee. The root mean square error (RMSE) between network-derived exposure estimates and personal DRI measurements for mobile employees was 0.15 mg/m³, 1 ppm, 82 ppb, and 3 dBA for PM, CO, O₃, and noise, respectively. Pearson correlation between network-derived exposure estimates and DRI measurements ranged from 0.39 (noise for mobile employees) to 0.75 (noise for stationary employees). Despite the error observed estimating personal exposure to occupational hazards it holds promise as an additional tool to be used with traditional personal sampling due to the ability to frequently and easily collect exposure information on many employees.

Metal-Organic Framework-Surface-Enhanced Infrared Absorption Platform Enables Simultaneous On-Chip Sensing of Greenhouse Gases

Simultaneous on-chip sensing of multiple greenhouse gases in a complex gas environment is highly desirable in industry, agriculture, and meteorology, but remains challenging due to their ultralow concentrations and mutual interference. Porous microstructure and extremely high surface areas in metal-organic frameworks (MOFs) provide both excellent adsorption selectivity and high gases affinity for multigas sensing. Herein, it is described that integrating MOFs into a multiresonant surface-enhanced infrared absorption (SEIRA) platform can overcome the shortcomings of poor selectivity in multigas sensing and enable simultaneous on-chip sensing of greenhouse gases with ultralow concentrations. The strategy leverages the near-field intensity enhancement (over 1500-fold) of multiresonant SEIRA technique and the outstanding gas selectivity and affinity of MOFs. It is experimentally demonstrated that the MOF-SEIRA platform achieves simultaneous on-chip sensing of CO2 and CH4 with fast response time (<60 s), high accuracy (CO2: 1.1%, CH4: 0.4%), small footprint (100 × 100 µm2), and excellent linearity in wide concentration range (0-2.5 × 104 ppm). Additionally, the excellent scalability to detect more gases is explored. This work opens up exciting possibilities for the implementation of all-in-one, real-time, and on-chip multigas detection as well as provides a valuable toolkit for greenhouse gas sensing applications.

Sources of error and variability in particulate matter sensor network measurements

The quality of mass concentration estimates from increasingly popular networks of low-cost particulate matter sensors depends on accurate conversion of sensor output (e.g., voltage) into gravimetric-equivalent mass concentration, typically using a calibration procedure. This study evaluates two important sources of variability that lead to error in estimating gravimetric-equivalent mass concentration: the temporal changes in sensor calibration and the spatial and temporal variability in gravimetric correction factors. A 40-node sensor network was deployed in a heavy vehicle manufacturing facility for 8 months. At a central location in the facility, particulate matter was continuously measured with three sensors of the network and a traditional, higher-cost photometer, determining the calibration slope and intercept needed to translate sensor output to photometric-equivalent mass concentration. Throughout the facility, during three intensive sampling campaigns, respirable mass concentrations were measured with gravimetric samplers and photometers to determine correction factors needed to adjust photometric-equivalent to gravimetric-equivalent mass concentration. Both field-determined sensor calibration slopes and intercepts were statistically different than those estimated in the laboratory (α = 0.05), emphasizing the importance of aerosol properties when converting voltage to photometric-equivalent mass concentration and the need for field calibration to determine slope. Evidence suggested the sensors' weekly field calibration slope decreased and intercept increased, indicating the sensors were deteriorating over time. The mean correction factor in the cutting and shot blasting area (2.9) was substantially and statistically lower than that in the machining and welding area (4.6; p = 0.01). Therefore, different correction factors should be determined near different occupational processes to accurately estimate particle mass concentrations.