Analysis, Modeling, and Simulation for Traffic Incident Management Applications
Future Improvements in TIM AMS Applications
Based on the current state of the practice, several areas are ripe for further development in Traffic Incident Analysis Modeling and Simulation (TIM AMS). We have developed the following problem statements that should be considered for further research.
Project statements that are based on the priority area topics identified in the practitioner survey include the following:
- Incident Data (Standards, collection procedures, coordination with transportation agencies, etc.) – Covered in Topic 1;
- Incident Performance Measures (metrics, presentation, formats/reporting) – Covered in Topic 1;
- Technical Methods for Predicting Incident Congestion Extent (delay, reliability) – Topic 2 covers estimating incident delay and its contribution to total delay;
- Technical Methods for Predicting Incident Duration – Covered in Topic 3; and
- Technical Methods for Predicting Incident Crashes – Secondary Crashes covered in Topic 4.
Additional Topics were identified based on a review of current practice and research.
Topic 1 – Data Management/Data Dictionary for Archiving the Data Needed to Compute Incident Performance Measures
This effort would cover data collection and processing so that all agencies are calculating the same measures, ensuring that agencies could be compared meaningfully. One of the key items is the ability to track changes in a single incident over its duration – blockages and response characteristics can change – and capturing this allows more detailed analysis of strategies and more realistic estimates of delay to be modeled. This effort could be extended to cover other forms of disruptions, primarily work zones.
Purpose: The aim is to ensure a consistent process of data collection and archiving in order to enhance analysis and comparison of data across agencies.
Impact on practice: A consistent dictionary of terms and data collection procedures will significantly enhance and ease the level of effort involved in conducting analysis using incident-related data. Practitioners also will be able to communicate with each other more effectively as key words and terminology will mean the same across agencies.
Action: A series of reports modeled in a manner similar to the FHWA Traffic Analysis Toolbox to develop terminologies, criteria to be used in collecting data, and guidelines for archiving the data will be effective in setting up guidelines for archiving incident data.
How topic was developed: This topic was developed based on responses from the TIM survey. Agencies surveyed identified incident data standards, collection, and coordination as one of the top five priorities for which technical guidance would be helpful.
Topic 2 – Linking Incident Management Strategies to Incident Model Inputs
The ATDM/HCM project will lay out a conceptual framework for evaluating the traffic performance benefits of various incident management strategies, however; this work is based on very limited data to calibrate and validate the specific parameters and sensitivities included in the framework. Additional work is required to add additional strategies to the framework and to calibrate and validate the specific models in the framework. This type of information is valuable for all levels of modeling, including the new work under SHRP 2 L08, which is not carrying the analysis back to this level.
The need to understand how actions affect the inputs to the modeling process is heightened by the emergence of ATDM strategies. Some of these strategies are relatively new (e.g., lane control) while some have a longer history (e.g., ramp metering and incident management). These strategies are implemented dynamically and in combinations. Determining the conditions under which the strategies are triggered is key to modeling their effect. That is, having a “control plan” that replicates how operators will deploy ATDM strategies is required for effective modeling.
Purpose: To improve the confidence of the profession in the ATDM/HCM tools for evaluating TIM strategies and to increase the variety of TIM strategies that can be evaluated using the ATDM/HCM tools.
Impact on practice: This research will enable agencies to better understand the performance benefits of implementing various levels of TIM strategies and be able to make better decisions as to investments in TIM.
Action: This topic may take advantage of the SHRP 2 research database to obtain data on traffic performance during incident conditions. Developing data on more advanced TIM strategies, not in the SHRP 2 database may require field data collection, preferably as “before/after” studies.
How topic was developed: This topic was developed based on relevant information gathered from ongoing research on ATDM and travel time reliability (SHRP 2).
Topic 3 – Real-Time Management and Prediction Methods
The ability to forecast an incident impact immediately after an incident occurs, given its location and the state of the network, is crucial to advanced operational control strategies. This study will develop methods to predict short-term impacts of a range of incident scenarios and develop recommended mitigation strategies of each scenario. The combined forecasting and mitigation models should be designed so they can be integrated into operational incident management programs.
Purpose: The aim is to give operators of incident management programs and traffic management centers (TMC) tools they can use to improve their response to incidents.
Impact on practice: This project will help develop clear procedures that operators of incident programs can apply to effectively respond to incidents. It also will ensure a more consistent response pattern for day-to-day operations of TMCs.
Action: This project should be conducted in two phases. The first phase should be a state-of-the-practice study on how incident management programs currently are predicting impacts of incidents and the steps they take in responding to incidents. The second phase should use historical data to develop the models that forecast incident impacts and develop mitigation strategies. The models developed in Phase 2 should be tested in real-time scenarios at TMCs.
How topic was developed: This topic was developed based on responses from the TIM survey. Incident prediction and prediction of incident duration were the two top items for technical guidance identified by agencies participating in the TIM survey that wanted technical guidance.
Topic 4 – Identify Key Factors that Cause Secondary Crashes
This effort would define a standard methodology for identifying secondary crashes based on identifying the key mechanisms that would cause secondary crashes. These include queue location at the time of crash, rubbernecking in the vicinity of the crash for both directions, and crashes involving emergency vehicles on their way to manage the primary incident.
Purpose: This method would abandon crude assumptions of where and when a queue is present and use travel time and volume data to identify the queue empirically.
Impact on practice: The improved and more credible process of identifying secondary crashes will make it easier to conduct analysis on the impact of TIM programs on secondary crashes and also assess the impacts of secondary crashes (delay, queuing etc.).
Action: This will involve extensive historical or real-time data at crash sites with secondary crashes. The data collection should involve the factors that are likely to contribute to the secondary crashes, geometric location of the crash and secondary crash, queue length, flow of traffic in opposite direction of travel (rubbernecking), emergency vehicles are a few. Microsimulation modeling could be used to validate the models developed.
How topic was developed: Agencies participating in the TIM survey identified the need for improved technical methods for predicting secondary incident crashes.
Topic 5 – Modeling Progressive Lane Opening
It is very common for the lane blockage characteristics of an incident to change over the course of its duration. This can go both ways: responders may close additional lanes to facilitate clearance and involved vehicles may be moved. Recently, the idea of systematically opening lanes as clearance progresses, rather than waiting to open all lanes at once, has gained acceptance as a TIM practice. How this practice can be modeled remains an open question, and data on what are reasonable values for the duration of each successive-lane opening need to be developed.
Purpose: Develop improved approach to implementing progressing-lane opening strategies.
Impact on practice: Practitioners will have improved tools to assess the impacts of progressive-lane opening strategies and will be able to better assess the performance of such implementations and whether it is having the intended impact of improving traffic flow.
Action: Given that some agencies already are implementing the practice an NCHRP Synthesis study is suggested to survey the agencies and find out their current practices and analysis methods. This could be followed up with a full NCHRP study.
How topic was developed: This topic was developed because the literature scan showed a few agencies have begun implementing progressive lane opening after incident response. A synthesis of current practice will be useful in assisting other agencies that may be interested in implementing this strategy.
Topic 6 – Update of HCM Sections Regarding Incident Management and Capacity Loss
The data presented in the HCM on capacity loss is based only a few older studies and is not sensitive to any incident management strategies. These values are in need of updating with the improved traffic and incident data now available.
Purpose: Provide practitioners with more credible information and factors for conducting incident analysis.
Impact on practice: The updated values will ensure analyses conducted with the HCM are more acceptable to all stakeholders and analysts will be more confident about their results.
Action: Given the TRB Highway Capacity and Quality of Service Committee is responsible for the HCM it will be best to develop an NCHRP problem statement to conduct the study to update the factors in the HCM.
How topic was developed: This update is needed because the Highway Capacity Manual is a key resource and practice reference for all transportation professionals. Having the Manual address Incident Management will be critical in ensuring techniques and best practices are adopted by the professional transportation community.
Topic 7 – Estimating Incident Delay and its Contribution to Total Delay
Past studies attempting to do this are somewhat limited and need to be extended. In addition to a methodology that could be applied on a case-by-case basis, simple predictive equations could be developed that relate incident delay to traffic and basic incident characteristics.
Purpose: Improve the state of the practice in quantifying and analysis of incident impacts.
Impact on practice: Ability to better quantify delay impacts due to incidents will create an opportunity to develop more targeted and effective measures to address these delays.
Action: This will work well as an NCHRP research project. A key issue when modeling incident delay – as well as other nonrecurring causes of congestion – is how to account for recurring congestion. On facilities where a significant amount of recurring congestion occurs, the effect of incidents on total congestion is complex. If a lane-blocking incident occurs upstream of a bottleneck, the incident will meter flow into the bottleneck, reducing or eliminating the congestion it causes. In other words, the congestion cause has moved from the bottleneck to the incident location, i.e., “delay migration” has occurred. It is true that the incident-caused delay can be greater than the bottleneck-caused delay (depending on the number of lanes blocked), but accounting for the recurring delay that would have been there without the metering effect of the incident needs to be addressed.
How topic was developed: Agencies surveyed indicated incident prediction and duration were top on their list of topics for which they wanted additional technical guidance. This topic was developed to help analyze the impact of incident characteristics (location, severity, and duration) on delay.
Topic 8 – Use of Simulation in Tabletop Exercises
This topic consists of customizing the user interfaces and analysis methods of one or more simulation models to facilitate their use by nonsimulation experts in the testing of TIM strategies in a tabletop exercise format.
Purpose: To provide operators with tools for self-training and development of cost-effective TIM strategies tailored to their agency’s capabilities, objectives, and the characteristics of the incidents on their facilities.
Impact on practice: This would result in a significant improvement in the day-to-day quality and effectiveness of agency responses to incidents. The agency also would have a greater depth of experienced personnel to place in charge of incident management. Agencies also would be able to use the tabletop exercises to explore various “what-if” scenarios.
Action: Candidate simulation tools that can be readily modified into a tabletop exercise format must be identified. Speed of operation, sensitivity to TIM strategies, adaptability to tabletop exercise, and ease of use by nonexperts would be primary considerations. One or more cooperative simulation model developers may be partnered with to develop and test the necessary user interface modifications to facilitate rapid execution by personnel more familiar with incident management than simulation. Finally, the resulting model(s) must be tested with operations personnel to verify their ease of use by nonexperts and their ability to accurately reflect real world results.
How topic was developed: This topic was developed based on the review of the example application of TIM AMS strategies. The review showed that due to limited tools for incident analysis, most studies involve postprocessing data from a combination of simulation tools, travel demand models, and sketch planning programs (HERS/IDAS). Developing customized user interfaces and postprocessors will make it easier for nonexperts of the tools to understand TIM applications.
Topic 9 – Framework Development for TIM Program Evaluation
Framework development for TIM program evaluation is necessary to assist operating agencies in correctly identifying and quantifying the costs and benefits of TIM programs. A structured framework also will ensure more consistent information reporting which will enable more collaboration and knowledge exchange and sharing between agencies.
Purpose: To identify the inputs required and appropriate tools and analytical procedures for accurately evaluating the costs and benefits of TIM programs.
Impact on practice: The accurate assessment of TIM costs and benefits will support the development and implementation of cost-effective TIM programs.
Action: The primary and secondary benefits and costs of TIM and the variables affecting them must be identified. An economic, traffic operations, safety, and environmental assessment framework must then be developed.
How topic was developed: This topic was developed based on responses from the survey.
Topic 10 – Expand Benefit/Cost Analysis to Include Secondary Impacts
This topic is an extension of the analysis framework discussed under the Topic 9. This topic expands primary B/C analysis to include secondary impacts of TIM programs, particularly the reduction in secondary crashes that can be a significant benefit of good TIM programs.
Purpose: To identify and quantify the secondary impacts of TIM programs for use in benefit/cost analyses.
Impact on practice: Significantly improved assessment of the benefits of TIM programs, and improved decision-making regarding investments in TIM programs by considering a full range of expected impacts is expected.
Action: The research would involve gathering data on the effects of TIM on secondary crashes, travel behavior, lost productivity, environment (emissions), insurance costs, etc. Data gathering would cover much longer periods to capture the likelihood of secondary accident occurrence. Models relating TIM strategies to secondary effects would be developed. A primer on the assessment of primary and secondary effects of TIM (such as the AASHTO Red Book – A Manual on User Benefit Analysis of Highway and Bus-Transit Improvements) would be produced to guide agencies in the benefit/cost assessment of TIM.
How topic was developed: This topic was developed because the TIM survey identified the need for improved technical methods for predicting secondary incidents and crashes. A more structured process is needed to include the secondary incidents in benefit/cost analysis.
Topic 11 – Resource Optimization for TIM Plans
Incidents are, to an extent, random in nature. Planning to respond to them involves an element of estimating the number of incidents that will occur, where they will occur, the type of incident, and the level of severity. Agencies will deploy and position response teams based on estimates of these variables (probably from historical information) and the amount of resources (personnel, vehicles, tow trucks, etc.) they have. The TIM objectives will largely determine the incident response team, subject to the agency budget constraints. For example, the number of required patrol units at a given freeway section can be determined by the desired response time. Even in the case where there are adequate budgets, the process of determining the optimal location to position resources in order to respond effectively is not trivial.
Purpose: To develop procedures, analytical tools, and best practices that agencies rely on to plan, deploy, and coordinate their traffic incident management teams so they can respond effectively and clear the highest number of incidents with the minimal amount of resources.
Impact on practice: A well-developed planning tool and a guidebook with best practices will significantly ease the burden and reduce the level of guesswork smaller agencies and those with limited budgets face in running TIM programs.
Action: A guidebook should be developed that includes an optimization tool and synthesis of best practice from agencies that have effective TIM programs.
The theoretical research needed to solve analytical resource allocation problems already exists in various branches of transportation literature and fields like industrial systems engineering. What is needed is the translation of the problem: the resources (service patrol staff and equipment, law enforcement agencies, first responders, etc.), constraints (limited budgets, area to be patrolled, communication staff) into an optimization system. The resource allocation tool developed should be developed in standard off-the-shelf commercial software like Microsoft Excel. A user interface should be included to make it easy to use, including GIS and mapping software to facilitate deployment.
The best practice guidelines should cover practices at agencies in large metropolitan areas, suburbs, and rural areas to make the content useful to the diverse set of agencies that might use it.
How topic was developed: This topic was based on some of the challenges identified during the development of the report. We found agencies are still struggling with basic problems like the most optimal location for dispatching a freeway service patrol truck, or whether to have the truck drive up and down a corridor or position it. Such problems have already been solved in several areas and a guidebook will bring the solutions together in one place for managers of TIM programs.
