Coordination is an objective of signal timing, and management of signal offsets is a critical part of that task. This paper presents a visualization tool called the offset adjustment diagram. This diagram is based on an analysis of cyclic flow profiles. Using the principle that traffic conditions under different offsets can be predicted by shifting vehicle arrival profiles relative to the distribution of green times, it is possible to determine the range of possible performance for an approach by sweeping through the possible offsets. This performance range is then used to provide context for the current performance, as well as the predicted change in performance under proposed offset adjustments. The diagram can thus serve to summarize corridor condition and proposed changes at a glance. The paper uses several case studies from the field to illustrate the use of the offset adjustment diagram, including an analysis of six corridors across a region, assistance with manual offset adjustments, and a comparison of alternative outcomes from prioritizing different directions during optimization.

Adaptive signal control is the subject of an increasing amount of research, as well as development and implementation. Most existing adaptive control systems achieve coordination by applying system control as a constraining layer on top of local control. Some researchers have suggested that, with the right local-control logic, coordination might be achieved as an dynamically emergent phenomenon without the need for a management layer. This paper explores the potential of a self-organizing signal control algorithm using a variety of performance measures. First, the initially reported algorithm performance is reproduced in an idealized environment; next, the algorithm is applied in a realistic road network to compare its performance against actuated-coordinated control, with and without pedestrian phases. Comparisons are made under (1) the same base volumes used to design the actuated-coordinated timing plan; and (2) a variant volume. Self-organizing control is found to be more flexible than coordinated control, and induces a tradeoff in performance among different movement types. Delay reductions of 38–56% are seen in an environment without pedestrian phases. However, with pedestrian phases in recall, self-organizing control performs worse (39% increase in delay) under base volumes, and achieves a weak benefit (6% reduction in delay) under the variant volume. Because of the large total delay reductions in some scenarios, the results show promise for future development.

 The town of Moab, Utah, has a unique traffic management challenge due to the combination of seasonal tourist traffic, heavy truck traffic, and high pedestrian volumes, making the Main Street corridor a strong candidate for an adaptive traffic control system (ATCS). Moab's remote location adds additional challenges in monitoring traffic in real time and maintaining signal timing plans. This paper describes how peer-to-peer (P2P) communication and user-definable control logic (UDCL) were utilized to implement a cost-effective ATCS in Moab, referred to as P2P adaptive control, using only the existing local controllers and detection. The adaptive control logic adjusts green time along the mainline based on detector inputs, while keeping the side streets at the minimum time needed for pedestrian service. The performance of the system was evaluated by comparing operations before and after implementation of the P2P adaptive control, using performance measures generated from high-resolution signal controller data. The P2P adaptive control increased the through bandwidth of the corridor and reduced the number of split failures, which is the number of phase occurrences with insufficient green. Future work includes adjustments to the algorithm to improve side-street service and expansion of the P2P adaptive control to additional signals expected to be constructed in the area.

The increasing connectivity in transportation infrastructure is driving a need for additional security in transportation systems.  In order to make security decisions in a budget constrained environment, the possible effect of a cyber-attack must be numerically characterized.  The amount of an effect is dependent on the level of access and the vehicular demand on the intersections being controlled.  The paper proposes a framework for better understand the levels of access and the impact that can be had in scenarios with varying demand.  Simulations are performed on a simplistic corridor to provide numerical examples of the possible effects.  The paper concludes that the possibility of some levels of cyber threat may be acceptable in locations where the traffic volumes would not be able to create an unmanageable queue.  The more intimate levels of access can cause serious safety concerns by modifying the settings of the traffic controller in ways that encourage red light running and accidents.  The proposed framework can be used by transportation professionals and cyber-security professionals to prioritize the actions to be taken to secure the infrastructure.

Early research on Hardware-in-the-Loop Simulation (HILS) systems mainly focused on reliable ways of replicating real-world controller logics in traffic simulation experiments. Since the mid-2000’s, the HILS systems have been replaced with Software-in-the-Loop Simulation (SILS) which are more affordable and efficient, and easier to install and use. However, the recent advances in controller’s data storage and handling have brought opportunities for development of high-resolution performance measures, which have revived the need for HILS systems. Unlike before, when controllers were used mostly to pass the data from detectors to the CPU and pass signal status back to signal displays, controllers are increasingly being used for storage and handling large amounts of high-resolution data. This research reports on development of the HILS lab for testing high-resolution controller event logs and the subsequent retrieval of selected signal performance measures. Controllers from six major US controller vendors have been integrated into an HILS lab and tested through microsimulation experiments. High-resolution event logs are recorded and transferred to a web server where the SPMs are retrieved. The process of logging simulated events is monitored to report differences between controller specifications. The end product is set of minimum event log standards, which are expected to be fulfilled by the controller vendors in future deployment of their controllers in the state of Florida.

This paper discusses strategies implemented by the District Department of Transportation (DDOT) to enhance the reliability of its signal system. The reliability measures discussed in this paper include deploying backup generators and installing uninterruptable power supply (UPS) battery backup systems (BBS) for critical intersections to minimize disruptions during power outages. The measures have yielded beneficial results. The benefits were apparent not only during severe weather events such as the 2012 derecho and Hurricane Sandy, but also under normal circumstances. The paper quantifies benefits by analyzing data on signal outage and run-time of generators and BBS. Each deployment has implications not only on traffic flow and safety, but also on city resources (a dark signal at a critical intersection typically requires a response from police or traffic control officers). The paper makes the business case for investing in systems that enhance system operations through increased reliability. The paper also discusses how big data analytics can help jurisdictions devise their strategy on technology selection and deployment.

The efficient and safe movement of traffic at signalized intersections is the primary objective of any signal phasing and timing plan. Accommodation of left turns is more critical due to the higher need for balancing operations and safety. The objective of this study is to develop models to estimate the safety impacts of the use of left-turn phasing schemes. The models are based on data from 200 intersections in urban areas in Kentucky. For each intersection, approaches with a left-turn lane were isolated and considered with their opposing through approach in order to examine the left-turn related crashes. This combination of movements is considered to be one of the most dangerous in terms of intersection safety. Hourly traffic volumes and crash data were used in the modeling approach along with the geometry of the intersection.  The models allow for the determination of the most effective type of left-turn signalization based on the specific characteristics of an intersection approach. The accompanying nomographs provide an improvement over the existing methods and warrants and allow for a systematic and quick evaluation of the left-turn phase to be selected. The models utilize the most common variables that are already known during the design phase and can be used to determine whether a permitted or protected-only phase will suit the intersection when considering safety performance.

The evaluation of traffic signal systems on an agency level can be of great importance for identifying problems, self-assessment, budgeting, creating the strategy for future steps, etc. The most famous similar effort of this type is National Traffic Signal Report Card (NTSRC) which is used as an evaluation methodology for agencies country-wide.  The main difference in the proposed methodology is that it steps away from qualitative evaluation and grading and presents a new set of procedures for implementation of quantitative, therefore more unbiased, evaluation methodology. The proposed methodology should enable the conduction of self-evaluation and comparison between different agencies in terms of agency management, traffic signal operations, signal timing practices, traffic monitoring, data collection and maintenance. For two agencies, the numerical and logical values of the answers are used in the evaluation process to obtain the preliminary results with the factor of confidence which are displayed in order to explain the evaluation process.  The proposed methodology shows potential, especially if the number of the available data types increase with the introduction of high-resolution data logging controllers into regular operations. By including those additional performance measures, the methodology could be used for tracking the results of operating the traffic signals by the government institutions or the private companies.

Abstract

Since traffic signal controllers have many different characteristics, it is very important to have traffic signal simulation with actual signal controllers. Therefore, CIDs (controller interface devices) have been an important research tool in the field of traffic signal control.This paper documents the development of a virtual controller interface device (VCID), which connects traffic signal controllers and micro-simulation such as VISSIM. Different from previous CIDs, a VCID provides communication between controllers and simulation through NTCIP (National Transportation Communications for ITS Protocol) without the need of an actual hardware device. Two case studies were conducted and the VCID was compared with the emulator-in-the-loop simulation (EILS), hardware-in-the-loop simulation (HILS), and software-in-the-loop simulation (SILS) using a same transportation network. Queue length, travel time, and other relevant measures were considered in the evaluation.  The results showed that VCID is not only an accurate and reliable research tool, but also has many obvious advantages compared to other tools. It does not depend on any hardware, is easy to operate and easy to carry. Because it communicates through NTCIP, it can be used remotely as long as internet access is available, thus providing a convenient learning and testing environment for those who do not own a physical signal control laboratory. 

Keywords: traffic simulation, EILS, HILS, SILS, VCID, controller interface device

The goal of traffic signal coordination is to provide smooth progression along a corridor in order to reduce vehicle delay and the number of stops. Generally, algorithms that produce signal coordination plans attempt to minimize a performance index, such as a weighed combination of delay and stops or maximize bandwidth. Many optimization models only consider the objective on the coordinated route and don’t account for the impacts on non-coordinated movements and other system-wide impact. In addition, improving a driver’s perception about the operation of the signals is another goal of coordination. A systematic investigation of traffic signal coordination on a system was conducted to analyze the benefits and impacts. The vehicle time spent in the traffic system was measured as the basis of the analysis. A new measure that captures the concept of smooth progression is introduced to reflect the driver’s perception of the signals. This study was conducted using a VISSIM simulation model and the Multi Modal Intelligent Traffic Signal System (MMITSS) where vehicle trajectory data is available for computing the performance measures.

In comparison with emulator-in-the-loop simulation (EILS) or software-in-the-loop simulation (SILS), Hardware-in-the-loop simulation (HILS) is considered the most realistic real-time simulation by factoring in the complexity of hardware signal controllers. In conducting signalized intersection HILS, controller Interface Devices (CIDs) had been necessary components. However, the use of CIDs often makes the simulation framework more redundant by adding an extra hardware component in the loop. CID also adds communication overhead between simulation software and the signal controller, which may reduce the feasibility of conducting HILS. Besides, modern controllers embedded with advanced and propriety control algorithm are hard to emulate by generic simulation software. To improve exiting HILS framework, the concept of CID-free HILS is proposed. By replacing CIDs with a software module based on National Transportation Communication for Intelligent Transportation System Protocol (NTCIP), HILS becomes more efficient with less hardware redundancy. In view of the interchangeability and interoperability of NTCIP, the proposed CID-free HILS is expected to expand the scope of simulation as well as improve the degree of realism of HILS. The proof-of-concept (POC) test demonstrates that the CID-free HILS framework can be successfully applied for various simulation-based evaluation studies such as adaptive signal control technologies and Connected Vehicle applications.

A spillover detection method for urban traffic networks is proposed using only signal timing and stop line detector data. The method uses measurements from a local detector, with data from downstream intersections used for validation. A new measure, termed a spillover index, is introduced that denotes the reduced traffic flow for each signal cycle caused by spillovers. The spillover index is compared with a predefined threshold to detect spillovers. The proposed method is tested and validated on a small simulated network and a busy arterial road in the city of Dublin, which is controlled by the SCATS adaptive traffic signal control system. The evolution of spillovers on an arterial road is studied using a Markov Chain process. Field data are used to determine the transition probabilities among all possible spillover states. A sensitivity analysis is performed to study the detection accuracy.

Arterial traffic signal systems in the U.S., predominantly, deploy multiple signal timing plans to account for daily variability of traffic demand. These types of traffic flow deviations should be anticipated when timing plans are designed and therefore, serviced satisfactorily. When traffic flow patterns are no longer predictable, a predetermined time-of-day (TOD) plan may no longer be the optimal one. This research aims to examine the signal timing optimality by applying a method similar to traffic responsive plan selection to automatically recognize the best timing plan suited to the current traffic conditions. The proposed method attempts to determine whether signal timing settings’ optimality could be effectively estimated when systematic, major approach detector counts are available. The study uses four months of field microwave detector data coupled with turning movement count data obtained over several days. The findings showed that TOD signal timing plans mainly depend on an adequate data collection which best describes a specific set of traffic conditions. Thus, the designed plan is as optimal as the related traffic information is reliable, while a problem arises in the case of limited availability and low quality data. New technologies are capable of collecting and storing massive amounts of data. Even if the granularity of collected data is low, the data can be utilized to improve traffic performance i.e. reduce corridor delay. This could be of particular importance to traffic agencies that have installed, or plan to install new field devices.

While research surrounding diverging diamond interchanges has existed for more than a decade, the effort to standardize signal timing of the interchange has only recently developed.  This research sought to determine the influence of crossover spacing and increased volume levels on DDI phasing scheme performance. A three-factor fully crossed experiment was conducted. Vistro and DBAT were used to optimize the split, cycle length, and offset of each of the 72 treatments. Data was collected through microsimulation using Vissim. Mean interchange delay and mean stops per vehicle were selected as measures the effectiveness. Pairwise comparisons were used to determine if a preferred scheme existed given 1) a set spacing and increased volume; 2) a set volume regardless of the spacing; and 3) a set spacing regardless of the increased volume. The data revealed that most often a two- or three-critical phasing scheme resulted in the lowest mean interchange delay and number of stops. Overall, the results provide an initial signal timing scheme for practitioners given a crossover spacing, an increased volume, or both. Future work includes exploration of low volumes and balanced interchange volumes and their impact on the four-critical movement phasing scheme as well as the impact of closely spaced adjacent intersections.

    Adaptive area traffic control (ATC) systems are relatively expensive to deploy, requiring the installation, calibration, and regular maintenance of physical sensors. Due to the high level of technical and financial resources required, it is not untypical to find minimally functioning ATC systems in developing countries. Taking Cebu City, Philippines, as an example, the city’s Sydney Coordinated Adaptive Traffic System (SCATS) was installed before 2000, and as of January 2015, fewer than 35% of detectors were still functioning.

    To address this challenge, the project team has designed a study to determine whether taxi company GPS data are sufficient to evaluate and improve traffic signal timing plans in resource-constrained environments. If this work is successful, we may reduce the number of physical sensors required to support these systems, thereby substantially lowering installation and maintenance costs.

    To this end, using taxi GPS data provided by regional taxi hailing app, the team designed and implemented methodologies for evaluating the performance of traffic signal timing plans and for deriving updated “dynamic-fixed” plans, whereby fixed plans (with periods based on observable congestion patterns, as opposed to just time of day) are regularly iterated, until optimization is reached. To date, the team has conducted three rounds of iterations to ensure the stability of the proposed signal timings. From the exploratory analysis results, the team’s algorithm has been proven to be capable of generating reasonable green splits, though the adjustment of cycle length needs to be taken into consideration in the future.

Most macroscopic models for arterial signal coordination consider a single class of vehicles and use the average free-flow speed to represent vehicle’s mobility. The simulated arrival platoon is often uniform in density but, in reality, platoon dispersion takes place. The simulated traffic flow may not be reliable and the offsets to optimize coordination between signals along an arterial may become problematic. Some models can consider heterogeneous speed/mobility but are only applicable for freeways. On the other hand, Multi-class Cell Transmission Model (M-CTM) can be implemented in a signalized network and takes into account a mixture of vehicle types. This work briefly present an M-CTM model for arterial traffic and compares arterial signal offset optimization from the conventional Cell Transmission Model (S-CTM) and SYNCHRO with timing based on M-CTM when heterogeneous-speed passenger cars are presented in the network. The vehicles are classified into two groups based on their free-flow speeds. M-CTM produces the platoon dispersion phenomena and the arriving flow at the downstream intersection is no longer uniform over time. The offset optimization for arterial signal timing is developed based on the M-CTM traffic flow model. The optimal offsets from M-CTM is obtained by minimizing the total delay function. Using microscopic simulation software VISSIM as an evaluation tool, M-CTM is demonstrated to yield more accurate and better signal timing than S-CTM, SYNCHRO, and the traditional ideal offset of signal coordination, by accounting for different vehicle speeds and platoon dispersion along a coordinated arterial.

Identification of Traffic condition in a motorway is essential for traffic management and control. GPS-enabled cell phones provide new opportunities for location-based services and traffic estimation. When traveling on-board vehicles, these phones can accurately provide position and velocity of the vehicle, and therefore can be used as probe traffic sensors. The In this paper, a methodology for the identification of the traffic state from GPS sensor data found from a single probe vehicle is presented. The methodology combines Hidden Markov Model (HMM) and Support Vector Machine (SVM) techniques to classify individual data points. Later, an algorithm is developed to identify stop and go segments in a trajectory. An application of the methodology for traffic state dependent sampling in a freeway network in Albany, NY shows encouraging results.

Real-time queue length information at signalized intersections is useful for both performance evaluation and signal optimization. Previous studies have successfully examined the use of high resolution event-based data to estimate real-time queue lengths. Based on identification of critical breakpoints, real-time queue lengths can be estimated by applying the commonly-used shockwave model. Although breakpoints can be accurately identified using lane-by-lane detection, few studies have investigated queue length estimation using single-channel detection, which is a common detection scheme for actuated signal control. In this study, a break point misidentification checking process and two input-output models (upstream-based and local-based) are proposed to address the over-estimation and short queue length estimation problems of breakpoint-based models. These procedures are integrated with a typical breakpoint-based model framework and queue-over-detector (QOD) identification process. The proposed framework was evaluated using field-collected event-based data along Speedway Boulevard in Tucson, Arizona. Significant improvements in maximum queue length estimates were achieved using the proposed method compared to the breakpoint-based model, with MAEs of 35.7 ft. and 105.6 ft., respectively.

High vehicle speeds are strongly associated with both a greater likelihood of crash occurrence and more serious pedestrian injury. This study evaluated the effectiveness of traffic signal progression as a speed management tool in three arterial corridors in the city of San Francisco. Analysis of “before” and “after” field data on traffic volumes and speeds were used in the evaluation supplemented with estimates of air pollutant vehicle emissions. The findings show that the implemented control measure is an effective low-cost method to reduce the average speeds at the selected corridors. The revised signal settings timings also resulted in emissions reduction. 

We present a red light running (RLR) analysis conducted at saturated intersections in the city of Mumbai, India, where the traffic is highly heterogeneous. Considering all vehicles, almost one in seventeen drivers are seen to be jumping red signals here. Unlike the RLR behaviour that has been hitherto reported from intersections elsewhere, a peculiarity observed here is that within a single red phase, there exist two distinguishable segments of RLR behaviour. We classify them into two regimes: Regime 1, just after the onset of red and Regime 2, just before the onset of next green. Around one-third of RLR events occur in Regime 1, and the rest in Regime 2. We fit different distributions on the time distribution of RLR events. The Kolmogorov-Smirnov (K-S) test suggests that exponential distribution fits best for RLR behaviours in Regime 1, and extreme value distribution fits for Regime 2, at all intersections. In addition to these two regimes, RLR at a lower rate is observed in the time period between these regimes, and normal distribution fits here. To analyse the causal factors of RLR behaviour in the two regimes, we develop models at a mesoscopic level, specific to vehicle-class and regime. While `red to green ratio' and presence of policing proves to be relevant factors affecting RLR in both the regimes, `relative time for which conflict area is free' affects RLR in Regime 2, but not in Regime 1.