A Descriptive Analysis on the Impact of COVID-19 Lockdowns on Road Traffic Incidents in Sydney, Australia

COVID-19 has had tremendous effects worldwide, resulting in large-scale death and upheaval. An abundance of studies have shown that traffic patterns have changed worldwide as working from home has become dominant, with many facilities, restaurants and retail services being closed due to the lockdown orders. With regards to road safety, there have been several studies on the reduction in fatalities and crash frequencies and increase in crash severity during the lockdown period. However, no scientific evidence has been reported on the impact of COVID-19 lockdowns on traffic incident duration, a key metric for crash management. It is also unclear from the existing literature whether the impacts on traffic incidents are consistent across multiple lockdowns. This paper analyses the impact of two different COVID-19 lockdowns in Sydney, Australia, on traffic incident duration and frequency. During the first (31 March–28 April 2020) and second (26 June–31 August 2021) lockdowns, the number of incidents fell by 50% and 60%, respectively, in comparison to the same periods in 2018 and 2019. The proportion of incidents involving towing increased significantly during both lockdowns. The mean duration of crashes increased by 16% during the first lockdown, but the change was less significant during the subsequent lockdown. Crashes involving diversions, emergency services and towing saw an increase in the mean duration by 67%, 16%, and 47%, respectively, during the first lockdown. However, this was not reflected in the 2021 data, with only major crashes seeing a significant increase, i.e., by 58%. There was also a noticeable shift in the location of incidents, with more incidents recorded in suburban areas, away from the central business area. Our findings suggest drastic changes in incident characteristics, and these changes should be considered by policymakers in promoting a safer and more sustainable transportation network in the future.

Dynamic prediction of traffic incident duration on urban expressways: a deep learning approach based on LSTM and MLP

Purpose Efficient traffic incident management is needed to alleviate the negative impact of traffic incidents. Accurate and reliable estimation of traffic incident duration is of great importance for traffic incident management. Previous studies have proposed models for traffic incident duration prediction; however, most of these studies focus on the total duration and could not update prediction results in real-time. From a traveler’s perspective, the relevant factor is the residual duration of the impact of the traffic incident. Besides, few (if any) studies have used dynamic traffic flow parameters in the prediction models. This paper aims to propose a framework to fill these gaps. Design/methodology/approach This paper proposes a framework based on the multi-layer perception (MLP) and long short-term memory (LSTM) model. The proposed methodology integrates traffic incident-related factors and real-time traffic flow parameters to predict the residual traffic incident duration. To validate the effectiveness of the framework, traffic incident data and traffic flow data from Shanghai Zhonghuan Expressway are used for modeling training and testing. Findings Results show that the model with 30-min time window and taking both traffic volume and speed as inputs performed best. The area under the curve values exceed 0.85 and the prediction accuracies exceed 0.75. These indicators demonstrated that the model is appropriate for this study context. The model provides new insights into traffic incident duration prediction. Research limitations/implications The incident samples applied by this study might not be enough and the variables are not abundant. The number of injuries and casualties, more detailed description of the incident location and other variables are expected to be used to characterize the traffic incident comprehensively. The framework needs to be further validated through a sufficiently large number of variables and locations. Practical implications The framework can help reduce the impacts of incidents on the safety of efficiency of road traffic once implemented in intelligent transport system and traffic management systems in future practical applications. Originality/value This study uses two artificial neural network methods, MLP and LSTM, to establish a framework aiming at providing accurate and time-efficient information on traffic incident duration in the future for transportation operators and travelers. This study will contribute to the deployment of emergency management and urban traffic navigation planning.

Dynamic and Interpretable Hazard-Based Models of Traffic Incident Durations

Understanding and predicting the duration or “return-to-normal” time of traffic incidents is important for system-level management and optimization of road transportation networks. Increasing real-time availability of multiple data sources characterizing the state of urban traffic networks, together with advances in machine learning offer the opportunity for new and improved approaches to this problem that go beyond static statistical analyses of incident duration. In this paper we consider two such improvements: dynamic update of incident duration predictions as new information about incidents becomes available and automated interpretation of the factors responsible for these predictions. For our use case, we take one year of incident data and traffic state time-series data from the M25 motorway in London. We use it to train models that predict the probability distribution of incident durations, utilizing both time-invariant and time-varying features of the data. The latter allow predictions to be updated as an incident progresses, and more information becomes available. For dynamic predictions, time-series features are fed into the Match-Net algorithm, a temporal convolutional hitting-time network, recently developed for dynamical survival analysis in clinical applications. The predictions are benchmarked against static regression models for survival analysis and against an established dynamic technique known as landmarking and found to perform favourably by several standard comparison measures. To provide interpretability, we utilize the concept of Shapley values recently developed in the domain of interpretable artificial intelligence to rank the features most relevant to the model predictions at different time horizons. For example, the time of day is always a significantly influential time-invariant feature, whereas the time-series features strongly influence predictions at 5 and 60-min horizons. Although we focus here on traffic incidents, the methodology we describe can be applied to many survival analysis problems where time-series data is to be combined with time-invariant features.

Assessing the Bus Bridging Effectiveness on the Operational Resilience of the Subway Service in Toronto

Unplanned rail disruptions result in substantial delays to passengers and severe effects on the economy of a large city like Toronto. While bus bridging has been a widely adopted method to replace the subway service in such events, its effect on the operational resilience of the subway service is less often studied. This study assesses the resilience of the subway network of Toronto employing an optimal bus bridging strategy. First, subway incidents are categorized based on their characteristics using K-mean clustering analysis. The incidents are then grouped based on the performance of optimal bus bridging plans. Classification and regression tree analysis is used for this task, employing two metrics: the total user delay and total number of shuttle buses under the optimal bridging scenario. Queueing and optimization models developed previously by the University of Toronto are used to determine and simulate the optimal bus bridging plans of a sample of incidents. The severity of unplanned disruptions is finally demonstrated using a severity scale and the effect of incident duration uncertainty is analyzed. The results show that along the congested city alignments, where the capacity of the roads and stations is limited, the bus bridging service is often insufficient to replace the train service. However, it could be a good alternative in uncongested subway segments where the available street capacity is relatively high, allowing large bus volumes to serve the corridor. This model is easily applicable to different rail systems and it could assess other systems to produce better bridging plans.

Predicting Incident Duration Based on Machine Learning Methods

Traffic incidents dont only cause various levels of traffic congestion but often contribute to traffic accidents and secondary accidents, resulting in substantial loss of life, economy, and productivity loss in terms of injuries and deaths, increased travel times and delays, and excessive consumption of energy and air pollution. Therefore, it is essential to accurately estimate the duration of the incident to mitigate these effects. Traffic management center incident logs and traffic sensors data from Eastbound Interstate 70 (I-70) in Missouri, United States collected during the period from January 2015 to January 2017, with a total of 352 incident records were used to develop incident duration estimation models. This paper investigated different machine learning (ML) methods for traffic incidents duration prediction. The attempted ML techniques include Support Vector Machine (SVM), Random Forest (RF), and Neural Network Multi-Layer Perceptron (MLP). Root mean squared error (RMSE) and Mean absolute error (MAE) were used to evaluate the performance of these models. The results showed that the performance of the models was comparable with SVM models slightly outperforms the RF, and MLP models in terms of MAE index, where MAE was 14.23 min for the best-performing SVM models. Whereas, in terms of the RMSE index, RF models slightly outperformed the other two models given RMSE of 18.91 min for the best-performing RF model. Index Terms— Incident Duration, Neural Network Multi-Layer Perceptron, Random Forest, Support Vector Machine.

Do Larger Sample Sizes Increase the Reliability of Traffic Incident Duration Models? A Case Study of East Tennessee Incidents

Incident duration models are often developed to assist incident management and traveler information dissemination. With recent advances in data collection and management, enormous achieved incident data are now available for incident model development. However, a large volume of data may present challenges to practitioners, such as data processing and computation. Besides, data that span multiple years may have inconsistency issues because of the data collection environments and procedures. A practical question may arise in the incident modeling community—Is that much data really necessary (“all-in”) to build models? If not, then how many data are necessary? To answer these questions, this study aims to investigate the relationship between the data sample sizes and the reliability of incident duration analysis models. This study proposed and demonstrated a sample size determination framework through a case study using data of over 47,000 incidents. This study estimated handfuls of hazard-based duration models with varying sample sizes. The relationships between sample size and model performance, along with estimate outcomes (i.e., coefficients and significance levels), were examined and visualized. The results showed that the variation of estimated coefficients decreases as the sample size increases, and becomes stabilized when the sample size reaches a critical threshold value. This critical threshold value may be the recommended sample size. The case study suggested a sample size of 6,500 to be enough for a reliable incident duration model. The critical value may vary significantly with different data and model specifications. More implications are discussed in the paper.