Traffic Noise Assessment Using Intelligent Acoustic Sensors (Traffic Ear) and Vehicle Telematics Data.

Here, we introduce Traffic Ear, an acoustic sensor pack that determines the engine noise of each passing vehicle without interrupting traffic flow. The device consists of an array of microphones combined with a computer vision camera. The class and speed of passing vehicles were estimated using sound wave analysis, image processing, and machine learning algorithms. We compared the traffic composition estimated with the Traffic Ear sensor with that recorded using an automatic number plate recognition (ANPR) camera and found a high level of agreement between the two approaches for determining the vehicle type and fuel, with uncertainties of 1-4%. We also developed a new bottom-up assessment approach that used the noise analysis provided by the Traffic Ear sensor along with the extensively detailed urban mobility maps that were produced using the geospatial and temporal mapping of urban mobility (GeoSTMUM) approach. It was applied to vehicles travelling on roads in the West Midlands region of the UK. The results showed that the reduction in traffic engine noise over the whole of the study road was over 8% during rush hours, while the weekday-weekend effect had a deterioration effect of almost half. Traffic noise factors (dB/m) on a per-vehicle basis were almost always higher on motorways compared the other roads studied.

Mapping urban mobility using vehicle telematics to understand driving behaviour.

Telematics data, primarily collected from on-board vehicle devices (OBDs), has been utilised in this study to generate a thorough understanding of driving behaviour. The urban case study area is the large metropolitan region of the West Midlands, UK, but the approach is generalizable and translatable to other global urban regions. The new approach of GeoSpatial and Temporal Mapping of Urban Mobility (GeoSTMUM) is used to convert telematics data into driving metrics, including the relative time the vehicle fleet spends idling, cruising, accelerating, and decelerating. The telematics data is also used to parameterize driving volatility and aggressiveness, which are key factors within road safety, which is a global issue. Two approaches to defining aggressive driving are applied and assessed, they are vehicle jerk (the second derivative of vehicle speed), and the profile of speed versus acceleration/deceleration. The telematics-based approach has a very high spatial resolution (15-150 m) and temporal resolution (2 h), which can be used to develop more accurate driving cycles. The approach allows for the determination of road segments with the highest potential for aggressive driving and highlights where additional safety measures could beneficially be adopted. Results highlight the strong correlation between vehicle road occupancy and aggressive driving.

Telematics data for geospatial and temporal mapping of urban mobility: Fuel consumption, and air pollutant and climate-forcing emissions of passenger cars.

In this study, we use the approach of geospatial and temporal (GeoST) mapping of urban mobility to evaluate the speed-time-acceleration profile (dynamic status) of passenger cars. We then use a pre-developed model, fleet composition and real-world emission factor (EF) datasets to translate vehicles dynamics status into real-urban fuel consumption (FC) and exhaustive (CO2 and NOx) emissions with high spatial (15 m) and temporal (2 h) resolutions. Road transport in the West Midlands, UK, for 2016 and 2018 is the spatial and temporal scope of this study. Our approach enables the analysis of the influence of factors such as road slope, non-rush/rush hour and weed days/weekends effects on the characteristics of the transport environment. The results show that real-urban NOx EFs reduced by more than 14 % for 2016-18. This can be attributed to the increasing contribution of Euro 6 vehicles by 63 %, and the increasing contribution of diesel vehicles by 13 %. However, the variations in the real-urban FC and CO2 EFs are less significant (±2 %). We found that the FC estimated for driving under the NEDC (National European Driving Cycle) is a qualified benchmark for evaluating real-urban FCs. Considering the role of road slope increases the estimated real-urban FC, and NOx, and CO2 EFs by a weighted average of 4.8 %, 3.9 %, and 3.0 %, respectively. Time of travel (non-rush/rush hour or weed days/weekends) has a profound effect on vehicle fuel consumption and related emissions, with EFs increasing in more free-flowing conditions.

Detecting high emitting vehicle subsets using emission remote sensing systems.

It is often assumed that a small proportion of a given vehicle fleet produces a disproportionate amount of air pollution emissions. If true, policy actions to target the highly polluting section of the fleet could lead to significant improvements in air quality. In this paper, high-emitter vehicle subsets are defined and their contributions to the total fleet emission are assessed. A new approach, using enrichment factor in cumulative Pareto analysis is proposed for detecting high emitter vehicle subsets within the vehicle fleet. A large dataset (over 94,000 remote-sensing measurements) from five UK-based EDAR (emission detecting and reporting system) field campaigns for the years 2016-17 is used as the test data. In addition to discussions about the high emitter screening criteria, the data analysis procedure and future issues of implementation are discussed. The results show different high emitter trends dependent on the pollutant investigated, and the vehicle type investigated. For example, the analysis indicates that 23 % and 51 % of petrol and diesel cars were responsible for 80 % of NO emissions within that subset of the fleet, respectively. Overall, the contributions of vehicles that account for 80 % of total fleet emissions usually reduce with EURO class improvement, with the subset fleet emissions becoming more homogenous. The high emitter constituent was more noticeable for pollutant PM compared with the other gaseous pollutants, and it was also more prominent for petrol cars when compared to diesel ones.

An integrated 3D CNN-GRU deep learning method for short-term prediction of PM2.5 concentration in urban environment.

This study proposes a new model for the spatiotemporal prediction of PM2.5 concentration at hourly and daily time intervals. It has been constructed on a combination of three-dimensional convolutional neural network and gated recurrent unit (3D CNN-GRU). The performance of the proposed model is boosted by learning spatial patterns from similar air quality (AQ) stations while maintaining long-term temporal dependencies with simultaneous learning and prediction for all stations over different time intervals. 3D CNN-GRU model was applied to air pollution observations, especially PM2.5 level, collected from several AQ stations across the city of Tehran, the capital of Iran, from 2016 to 2019. It could achieve promising results compared to the methods such as LSTM, GRU, ANN, SVR, and ARIMA, which are recently introduced in the literature; it estimates 84% (R2 = 0.84) and 78% (R2 = 0.78) of PM2.5 concentration variations for the next hour and the following day, respectively.

Impacts of emergency health protection measures upon air quality, traffic and public health: evidence from Oxford, UK.

Emergency responses to the COVID-19 pandemic led to major changes in travel behaviours and economic activities in 2020. Machine learning provides a reliable approach for assessing the contribution of these changes to air quality. This study investigates impacts of health protection measures upon air pollution and traffic emissions and estimates health and economic impacts arising from these changes during two national 'lockdown' periods in Oxford, UK. Air quality improvements were most marked during the first lockdown with reductions in observed NO2 concentrations of 38% (SD ± 24.0%) at roadside and 17% (SD ± 5.4%) at urban background locations. Observed changes in PM2.5, PM10 and O3 concentrations were not significant during first or second lockdown. Deweathering and detrending analyses revealed a 22% (SD ± 4.4%) reduction in roadside NO2 and 2% (SD ± 7.1%) at urban background with no significant changes in the second lockdown. Deweathered-detrended PM2.5 and O3 concentration changes were not significant, but PM10 increased in the second lockdown only. City centre traffic volume reduced by 69% and 38% in the first and second lockdown periods. Buses and passenger cars were the major contributors to NO2 emissions, with relative reductions of 56% and 77% respectively during the first lockdown, and less pronounced changes in the second lockdown. While car and bus NO2 emissions decreased during both lockdown periods, the overall contribution from buses increased relative to cars in the second lockdown. Sustained NO2 emissions reduction consistent with the first lockdown could prevent 48 lost life-years among the city population, with economic benefits of up to £2.5 million. Our findings highlight the critical importance of decoupling emissions changes from meteorological influences to avoid overestimation of lockdown impacts and indicate targeted emissions control measures will be the most effective strategy for achieving air quality and public health benefits in this setting.

Real-world assessment of vehicle air pollutant emissions subset by vehicle type, fuel and EURO class: New findings from the recent UK EDAR field campaigns, and implications for emissions restricted zones.

This paper reports upon and analyses vehicle emissions measured by the Emissions Detecting and Reporting (EDAR) system, a Vehicle Emissions Remote Sensing System (VERSS) type device, used in five UK based field campaigns in 2016 and 2017. In total 94,940 measurements were made of 75,622 individual vehicles during the five campaigns. The measurements are subset into vehicle type (bus, car, HGV, minibus, motorcycle, other, plant, taxi, van, and unknown), fuel type for car (petrol and diesel), and EURO class, and particulate matter (PM), nitric oxide (NO) and nitrogen dioxide (NO2) are reported. In terms of recent EURO class emission trends, NO and NOx emissions decrease from EURO 5 to EURO 6 for nearly all vehicle categories. Interestingly, taxis show a marked increase in NO2 emissions from EURO 5 to EURO 6. Perhaps most concerningly is a marked increase in PM emissions from EURO 5 to EURO 6 for HGVs. Another noteworthy observation was that vans, buses and HGVs of unknown EURO class were often the dirtiest vehicles in their classes, suggesting that where counts of such vehicles are high, they will likely make a significant contribution to local emissions. Using Vehicle Specific Power (VSP) weighting we provide an indication of the magnitude of the on-site VERSS bias and also a closer estimate of the regulatory test/on-road emissions differences. Finally, a new 'EURO Updating Potential' (EUP) factor is introduced, to assess the effect of a range of air pollutant emissions restricted zones either currently in use or marked for future introduction. In particular, the effects of the London based Low Emission Zone (LEZ) and Ultra-Low Emissions Zone (ULEZ), and the proposed Birmingham based Clean Air Zone (CAZ) are estimated. With the current vehicle fleet, the impacts of the ULEZ and CAZ will be far more significant than the LEZ, which was introduced in 2008.