Survey of Work Zone Practices and Validation of Highway Capacity Manual Work Zone Capacity Model

Capacity at work zones is one of the major factors affecting queueing at work zones. Different states within the United States use their own methodology in determining work zone capacities and when to implement lane closures at work zones. The objective of this study was two-fold: first, to provide a synthesis of work zone lane closure procedures practiced by the various Departments of Transportation (DOTs) nationwide; and secondly, to validate the Highway Capacity Manual 6th edition’s (HCM 6) work zone capacity model using field-collected data in the state of Louisiana. The first objective was met by administering a survey to DOTs nationwide. The survey revealed that half of the states that responded to the survey require minimum capacity for short-term work zone lane closures, with minimum capacity ranging from 1100 to 1900 passenger cars per hour per lane. In addition, most of the states reported implementing consistent policies across various district offices. The survey findings provide a good source of information on queue analysis and work zone lane closure policies adopted across different DOTs. The second objective was met by collecting traffic flow data from 10 work zone sites within the state of Louisiana and validating the capacity model in the HCM 6. Results showed the HCM 6 model slightly overestimating the average field-observed capacity by 6%. In the absence of local data, the HCM 6 model provides a great tool to estimate work zone capacities in Louisiana.

The Impact Of Lane Closures On Congestion And Reliability: A Case Study Of Toronto's Gardiner Expressway

Usage of big data with before-after methods of analysis makes it possible to evaluate the effect of major transport investments on system performance. In employing before-after methods to investigate the impact of lane closures on congestion and travel reliability, changes and trade-offs in performance indicators are quantified and policy action effectiveness is evaluated. This is illustrated through a case study of two separate lane closure interventions on the Gardiner Expressway in Toronto, Ontario. Models using a regression framework were developed for the pre-, peri-, and post-closure test periods of the first intervention and pre- and peri-closure periods of the second intervention. Results suggest the impacts of policy actions on system performance are strong, and that congestion and travel reliability counterintuitively move in different directions. Reduced demand effects are observed, prompting discussion on how highways and congestion should be managed and whether or not municipalities should add capacity to regional assets.

The Impact Of Lane Closures On Congestion And Reliability: A Case Study Of Toronto's Gardiner Expressway

Usage of big data with before-after methods of analysis makes it possible to evaluate the effect of major transport investments on system performance. In employing before-after methods to investigate the impact of lane closures on congestion and travel reliability, changes and trade-offs in performance indicators are quantified and policy action effectiveness is evaluated. This is illustrated through a case study of two separate lane closure interventions on the Gardiner Expressway in Toronto, Ontario. Models using a regression framework were developed for the pre-, peri-, and post-closure test periods of the first intervention and pre- and peri-closure periods of the second intervention. Results suggest the impacts of policy actions on system performance are strong, and that congestion and travel reliability counterintuitively move in different directions. Reduced demand effects are observed, prompting discussion on how highways and congestion should be managed and whether or not municipalities should add capacity to regional assets.

Development of Breakdown Probability Models and Heavy Vehicle Passenger Car Equivalents for Rural Freeway Work Zones

With nearly nine million lane-miles of public roadway and an economy driven by the automobile, interruptions to normal traffic operations for construction and maintenance are inevitable in the U.S.A., but the substantial safety and mobility impacts associated with queueing at freeway lane closures are mitigable. The current freeway work zone capacity methodology in the 6th edition of the Highway Capacity Manual is a vast improvement over historical guidance but still approaches the issue differently than research suggests agencies and practitioners should. Namely, a capacity defined by the mean queue discharge rate is deterministic and fails to account for the stochastic nature of traffic flow and breakdown. These core issues were addressed in this research by developing a methodology for obtaining probabilistic estimates of rural freeway work zone capacity from simulated data in PTV Vissim. Results for a two-to-one lane closure were presented as a series of breakdown probability distributions to demonstrate the viability of this methodology. The data indicated that the impact of trucks on freeway capacity is exacerbated in the presence of lane closures and led to the development of work zone capacity-based passenger car equivalents. Such a procedure may be extended to freeway facilities exhibiting different geometric, traffic, and environmental characteristics and utilized by agencies to make data-driven, risk tolerance-based planning, design, and operations decisions at freeway work zones.

Comparison of Three Traffic Management Plans Showing Shadow and Police Vehicle Effects on Driver Behavior at Highway Single Lane Closures

Lane closures on multi-lane roads require drivers to transition safely to an open lane before passing the worksite. To reduce worker and driver injury risk, truck-mounted attenuators (TMAs) are often used to prevent vehicle work zone intrusions and reduce the severity of collisions. To maximize the efficiency and effectiveness of TMA use, it is necessary to determine how and when they should be deployed as well as the best supporting measures. The current research focuses on the effects of different traffic management plans (TMPs) on driver behavior. Three TMPs at night time highway work zones were examined: ( 1 ) two tail vehicles in the advance warning area, ( 2 ) three tail vehicles in the advance warning area, and ( 3 ) addition of a marked police car with flashing lights in the buffer area downstream of the TMA. Driver response to the different TMPs was assessed by measuring vehicle speeds at three points in the traffic management area and observing lane change and merging behaviors on the approach to the TMA. Analysis showed a positive effect of police presence in the buffer area on driver behavior: TMP3 produced a reduction of 8.4%–12.9% in proportions of vehicles exceeding the speed limit by at least 5 km/h when passing the TMA. TMP3 also appeared to produce a positive effect on merging behavior compared with the other layouts. Use of a third tail vehicle in the advance warning area was not found to produce any additional safety benefit and may have a detrimental effect.

Machine Learning Approach to Forecast Work Zone Mobility using Probe Vehicle Data

The aim of deploying intelligent transportation systems (ITS) is often to help engineers and operators identify traffic congestion. The future of ITS-based traffic management is the prediction of traffic conditions using ubiquitous data sources. There are currently well-developed prediction models for recurrent traffic congestion such as during peak hour. However, there is a need to predict traffic congestion resulting from non-recurring events such as highway lane closures. As agencies begin to understand the value of collecting work zone data, rich data sets will emerge consisting of historical work zone information. In the era of big data, rich mobility data sources are becoming available that enable the application of machine learning to predict mobility for work zones. The purpose of this study is to utilize historical lane closure information with supervised machine learning algorithms to forecast spatio-temporal mobility for future lane closures. Various traffic data sources were collected from 1,160 work zones on Michigan interstates between 2014 and 2017. This study uses probe vehicle data to retrieve a mobility profile for these historical observations, and uses these profiles to apply random forest, XGBoost, and artificial neural network (ANN) classification algorithms. The mobility prediction results showed that the ANN model outperformed the other models by reaching up to 85% accuracy. The objective of this research was to show that machine learning algorithms can be used to capture patterns for non-recurrent traffic congestion even when hourly traffic volume is not available.

Estimation of the Effect of Rain and Incidents on Freeway Capacity and Free-Flow Speed

This research aims to examine the effect of incidents with lane closures and adverse weather conditions (medium to heavy rain intensity) on capacity and free-flow speed (FFS) of freeway segments. Data were collected from multiple freeway segments located in the Kansas City, U.S., metro area from 2014 to 2018. The capacity and FFS were measured for two-lane, three-lane, and four-lane freeways under four conditions: ( 1 ) base conditions, ( 2 ) adverse weather only, ( 3 ) incidents only, and ( 4 ) adverse-weather-and-incidents. Capacity adjustment factors (CAF), and speed adjustment factors (SAF) were established to identify the remaining capacity or the FFS reduction during an incident or adverse weather conditions. The findings indicated that medium to heavy rain resulted in a 5% reduction in FFS at three-lane sites which is consistent with the adjustment factors shown in the Highway Capacity Manual 6th edition (HCM6); however, rain was not found to have a significant impact on freeway capacity. It was also found that incidents leading to one-lane closures reduced capacities by 30%, 17%, and 17% at two-lane, three-lane, and four-lane sites, respectively. Incidents were also found to reduce FFS by approximately 5%–10%, possibly because of “rubbernecking.” Adjustment factors that capture the combined effect of incidents and rain on FFS and capacity are also presented.