Improving Traffic Signal Management and Operations: A Basic Service Model

V. Implementing a Basic Service Concept: The Traffic Signal Management Plan

Previously, this report presented the concept of good basic service as a reflection of agency archetypes that most effectively focused their resources, however plentiful or scarce, on their most important objectives. The interviews revealed that even the best agencies have difficulty articulating and maintaining that focus. This section will outline key strategies for doing so, and an outline of how to embody those strategies in a Traffic Signal Management Plan.

The following principles provide the basis for an agency to develop an objective-based traffic signal management program:

Clarity of objectives.
Attainable performance evaluation linked to those objectives.
Standards of performance (intended, for example, to score well on the Signal Timing Report Card) based on objectives rather than agency activities, especially arbitrary frequency of activities.
Resource requirements based on objectives rather than industry norms.
Clear and consistent communication with policy makers and elected officials.
Systems engineering in thought rather than merely in deed.

These strategies are discussed in the sections that follow.

Clarity of Objectives

As previously described, a good basic service concept is based on a simply stated objective. Drivers have simple overall expectations:

I want to drive to my destination at my desired speed with the minimum of attention. In the absence of achieving that goal, I want to be treated fairly and predictably so that I can plan my day with the minimum of uncertainty.

This leads to a simple objective for agencies to articulate:

We will do our best to avoid making drivers stop, but when we must make them stop, we will delay them as little as possible, within the context of safe operation.

As previously described, these objectives lead to a few high-level strategies for the agency to promote, including:

Field infrastructure reliability.
ignal timing that minimizes and balances congestion.
ignal timing that promotes smooth flow. This objective is an amalgamation of more narrowly cast objectives, such as delay, stops, and so on, but those objectives are evaluated in light of the broader objective of smooth flow.
Signal operation that corresponds to conditions predictably and consistently.
Versatile signal timing that provides broad-banded solutions rather than being narrowly optimized for a specific condition. (See Appendix 1 for further discussion.)

Attainable Performance Measures and Standards of Performance

If the strategies listed above are relevant to achieving the goal of fulfilling motorist expectations, the performance measures should be related to those strategies. The interviews revealed that even the best agencies find it difficult to measure their performance (other than in terms of citizen complaint calls) and stay focused on their objectives. Too often, agencies focus on traffic performance objectives such as delay and stops that are based on aggregate measures that are not directly perceivable by motorists, and too often the achievement of good maintenance is hidden by the effects of traffic growth with respect to network capacity outside the control of traffic signal operations staff.

Thus, an effective operational plan will define how the agency has implemented and measured the progress of those strategies.

Field Infrastructure Reliability.

The broadest measure of the reliability of the traffic infrastructure is:

The percentage of time motorists see what traffic agencies intend for them to see.

There are two general approaches for accomplishing this objective that emerge from the discussion of agency archetypes: By design and by maintenance response. Maintenance response includes emergency and preventive maintenance, and agencies have difficulty establishing the latter based on their resources being consumed by the former.

Agencies that have successfully implemented preventive maintenance programs have often built new programs, with new technicians and new equipment, based on clear productivity objectives. For example, the City of Austin, Texas established a preventive maintenance program in the mid-1980's that resulted in the addition of two technicians, each with a small service van equipped with a short telescopic bucket. These technicians were entirely devoted to preventive maintenance, as a program for which funding was sought separately and distinctively by the agency staff from the local policy makers and elected officials. The program was based on the principle that the agency would not have to expand its maintenance force in future years to keep up with additional demand for emergency maintenance because of the effects of preventive maintenance.

Agency staff measured and reported the effectiveness of that program in terms directly linked to the objective of the program: How many traffic signals received preventive maintenance during the year? A performance expectation for the technicians was established at four intersections per shift for preventive maintenance, and the technicians worked a night shift to avoid congested periods. What the agency did not do was close the loop to analyze whether average emergency maintenance declined as a result of improved preventive maintenance, and as a result lost focus on the program after several years and folded the new technicians into the regular technician pool. As with the Report Card, performing the activity alone is not enough. With maintained focus, the agency would likely have been able to report a decrease in the number and severity of emergency maintenance calls.

The same is true on the system side for coordinated systems. In order to increase the percentage of time the motorists see what the agency intends them to see, the operation must actually be observed. The interviews revealed that agencies did not observe the operation of their signal systems routinely to make sure they were running as intended. To some extent, this forms one of the fundamental purposes of a traffic signal system, which is to monitor the local infrastructure and report faults when they occur. Thus, the need for field observation can be minimized by a design solution rather than by merely increasing the maintenance function. But requirements leading to such design must be clearly articulated.

Signal Timing

Measuring traffic performance can be related to two very different questions:

Has the signal operation degraded to the point where it needs to be revised?
Is the signal operation good?

The first question has usually been the result of an arbitrary time limit, and the Report Card evaluates effectiveness here on the basis of how often an agency revises its signal timings. But an activity-based solution is not necessarily an effective use of limited resources.

In an ongoing project considering this question, one result has been that signal timing designed to accommodate heavier traffic (short of congestion) will also accommodate lighter traffic, with little impact on travel time. That work defines the solution space for signal timing as ranging from nearly no traffic to the point of congestion in a typical day (which may be in the future). Considering that agencies usually provide effective signal timing with four or five coordination patterns for this comprehensive range of conditions, the research pointed out that trends over time could be accommodated by providing signal timing plans designed to cover the solution space at the outset. (See the discussion on versatility in Appendix 1.)

The thornier question is evaluating given signal timings in absolute terms, especially in congested networks. The interviews revealed a very simple metric for determining whether signal timings seemed reasonable and equitable to motorists: Citizen complaints. Thus, one means of measuring improvements in this objective is by summarizing citizen complaint activity. Raw numbers are probably not helpful, however, because notice in the press can cause a public reaction that would not otherwise have occurred. These anomalies should be measured and discounted. Most agencies now track citizen complaints. One of the interviewers, for example, reported that all complaints were processed by a central operator at the agency, and tracked to ensure that they are resolved timely. A reduction in the number of complaints per signal, or the failure of complaints to keep pace with growth, would indicate progress in this area. The prevention of a sudden and sustained increase in complaints would indicate that the agency has not slipped in this measure.

By measuring signal timing in this way, agencies can avoid the enormous task of attempting to measure traffic performance comprehensively. But for agencies that have the resources, there are ways to collect relatively simple performance data using travel time data collection, as suggested by some of the archetypes, using a pair of technicians (one for the morning period and one for the afternoon period) for each 200 or so arterial systems that are evaluated on an annual basis. These technicians can collect peak-period travel time data in half a dozen runs on an arterial, from Monday afternoon to Friday morning, keeping track of season-over-season or year-over-year trends.

Objective-Based Resource Allocation

With a basic service concept based on keeping traffic moving, many of the activities undertaken by agencies may be seen as tangential to their objectives. In developing a traffic signal management plan, each agency should consider how each activity actually contributes to achieving their objectives. A commitment to the operation on the street was reported as critical during the interviews, and the agency archetypes most committed to street performance can be seen to achieve the best results.

For example, an operations plan should include a description of how the agency will monitor signal operation to make sure it is operating as designed, and a description of how the agency will observe that traffic flow is still being accommodated. For agencies with severely limited resources, simple observation may be the only choice, but even that is often undone even by agencies that don't face such severe limitations. While agencies usually monitor their response to citizen complaints, they should also provide a mechanism for responding to complaints generated by their own staff. For example, the City of San Antonio published time-space diagrams for all coordinated arterial streets in a notebook, and provided those (with training in their use) to all traffic-department employees who spend their day on the street. Their instructions were to pay attention to flow anomalies, such as unexpected queuing or unexpected repeated stops. This is one simple and inexpensive mechanism for monitoring system operation. Approaches like this are particularly useful for state agencies that have widely dispersed small-scale systems that often don't provide communication to the traffic operations office.

The operations plan should also include a description of the mechanism by which these observations are followed up with review by operations staff. It is not the recommendation to impose a cumbersome tracking program, but observations from staff should be afforded at least as much accountability as citizen complaints, and this is often not the case.

The focus of the operations plan should be on results, not on activities. A system that provides skilled qualitative review of every coordinated system in the jurisdiction of the agency every year is preferable to a system that provides, say, microsimulation of only a few systems receiving public or political attention. The difference between these is the same as the difference between preventive and emergency maintenance. The resource-intensive analysis methods, including those that require detailed data collection and analysis, should be reserved for problems that cannot be solved by simple skilled observation.

But agencies should also include in their operations plan a commitment to the basic design of the signal timing for the system. A series of qualitative revisions to operation may be counterproductive if it undermines the basic operational concept. For example, a series of signal timing adjustments may ruin progression without the technician making the adjustment realizing it. All such adjustments should be vetted in light of the original intent of the operation, and if an adjustment is needed that can't be accommodated within that original concept, the agency should consider developing new timings.

Many such approaches are possible, but the important recommendation is that agencies don't ignore the responsibility to allocate resources based on operational objectives rather than merely on activities.

Clear Communication Up the Line

Most traffic signal practitioners prefer to work anonymously within their agencies. This attitude is a natural outgrowth of the negative experiences that often accompany such interaction. In many agencies, traffic signal staff are precluded from direct contact with elected officials, and frequently are put between the rock of citizen complaints and the hard place of agency upper management and public-relations staff that are incentivized to please politicians. This uncomfortable position often leads to a cynicism that can undermine the effectiveness of the agency more deeply than any budget shortfall. One common outcome is that responding to citizen complaints and requests for information from elected officials is often delegated to non-professional staff who are technically skilled but perhaps not able present the information consistently with agency goals and strategies.

An operations plan should include the description of a mechanism by which the progress of programs, including routine operational programs, is reported to agency management, policy makers, elected officials, and the public. The mechanism should allow both ad hoc and systematic reporting, and all the reporting should be based on operational objectives rather than on agency convenience.

These communications need not be excessively formal or highly produced. The City of Arlington, Texas developed a means of communicating to management in the form of a technical memorandum called an "Informal Report to Mayor and City Council." These should be short and descriptive, rigorously factual, and closely related to the interests of the motoring public. One of them, for example, reported on the results of an agency-wide signal retiming effort conducted to update timings five years after a signal system had been installed. As with all the informal reports, the organization of the report included a background, discussion, and conclusion. Improvements were based on actual measurement of travel times, using a floating car method. Analysis related these travel times to average speed, stops, delay, and emissions, but travel times received top billing. The report provided only a summary of the entire network, and was less than two pages. During each year, the agency developed dozens of these reports, and they were the basis for winning the ITE Urban Traffic Engineering Achievement Award.

The City of San Antonio was more limited in resources by far, but still established basic principles for communicating outside the operations staff. These were:

The story told to citizens is the most important story in the agency. If it is ineffective in achieving citizen support and satisfaction, it will also not achieve policy and political satisfaction. Consistent with this importance, all citizen complaints concerning signal operation were not delegated, but handled directly by the professional in charge of the program. This had the dual benefit of telling the story most effectively, and making sure that the creator of the policy learned how to express it most effectively.
Requests for information from top management and elected officials must be responded to meaningfully within seven calendar days. This included those requests that required data collection, such as signal warrant studies.
Information presented to upper management would be presented with a high standard of professionalism, empowering and enabling those upper managers to represent that information to policy makers and elected officials more effectively. For example, the before-and-after travel-time studies associated with new traffic signal timings were summarized in a professionally produced report that extended the memorandum concept mentioned above to a report that could be distributed to the public.

An operations plan should include principles for communicating outside the agency, with special emphasis given to reporting successful completion of routine and special projects, and to making it the direct responsibility of the person in charge of the program to articulate it to the public. Those principles should be directly related to the objectives of the program. For example, a new signal system might be built to improve the management of field infrastructure, in which case the report should be related to that objective, e.g. the number of citizen complaints that were responded to without a field visit because of the system, the number of faults that were discovered and corrected before the public complained, the increased number of intersections updated with new and improved timings as a result of more efficient database management, and so on. Signal timing projects, on the other hand, should show the benefits of new timings in terms most relevant to motorists: Minimize travel time and travel time unreliability, which is usually interpreted as unpredictable and inequitable operation by motorists.

Meaningful Systems Engineering

An operations plan should include a requirement that all projects follow a meaningful systems engineering approach. This requirement should apply to all activities defined as projects, with an objective, a strategy, a tactical implementation, and a result.

This is particularly true for the replacement of signal systems, which are the most expensive and perceived to be the least successful of traffic signal management projects.

Most signal systems are justified based on traffic performance and safety. Yet these improvements result from signal timing improvements, not signal systems, and can be achieved through such approaches as time-based coordination.

Most agencies contemplating a signal system, however, have real needs not served by time-based coordination. Those needs can be summarized simply: Manage field infrastructure.

Systems engineering is a process by which real needs are identified first, and then the project is formulated and tested to confirm that those needs are addressed. As an established branch of systems development, systems engineering is well documented in wide-ranging literature (in addition to being a requirement for Federally funded projects, see here: www.ops.fhwa.dot.gov/int_its_deployment/sys_eng.htm and generally follows this process:

Document what the agency does using the proposed system. (This is not what the system does, which will be developed later, but what the people do that the system must support. As such, it doesn't describe the system at all; it describes the agency.) These activities are then broken down into a list of needs, called user needs. The documentation of this step is called a Concept of Operations.
For each need, define a requirement or series of requirements that the system must fulfill in order to address that need. These requirements form the Requirements Document and show explicitly how the requirements trace back, point by point, to the user needs.
Design the system such that all the requirements are fulfilled. This is the first step where the design of the system is contemplated, and this is the first description of the system itself. The design includes the normal technology analysis, design documents, and ultimately plans and specifications. Every design feature must be justified by playing a role in fulfilling at least one requirement, and every requirement must be fulfilled. One of the design documents is the Requirements Traceability Matrix, which explicitly documents how the design fulfills the requirements.
Test system components against the design. This is called Verification.
Test the system's ability to support the activities defined in the Concept of Operations. This is called Validation.

With that outline, a signal system project Concept of Operations will include a plain description of what operations staff will do. For example, when a citizen calls the practitioner with a complaint about the operation of a signal, the practitioner needs to be able to observe the operation of that signal in real time to evaluate the complaint. Many causes of the complaint can be observed by watching the signal operate through the system. Thus, the agency's activity is to respond to a citizen complaint by first observing the operation of the signal in real time, checking that:

The signal is not on conflict flash
The signal is in coordination
The pattern is correct
The controller time of day is correct
The timing has not been modified
The detectors are cycling normally
The signal is cycling normally

Further, the practitioner may benefit from seeing the traffic demand at the intersection using some form of traffic monitoring.

Based on what the observation reveals, the practitioner may dismiss the complaint, determine that a malfunction has occurred and refer it to maintenance technicians (or repair it through the system), or determine that the signal timing needs closer observation for possible revision.

The forgoing description of the agency's activity is the critical basis for determining the need for a traffic signal system. They provide a clear basis for requirements that will be used to evaluate the design of the system, such as the requirement to provide a real-time display, the requirement to include detector actuations in that display, the requirement to build a communications network that can keep up with that real-time display, and so on. Those activities should therefore be the subject of a proper Concept of Operations. The requirement, for example, that the system software include a real-time display can be traced directly back to this activity.

Most agency-developed systems engineering approaches focus on the system at the outset, rather than on the activities of the agency, with the result that the system eventually does not fully support those activities, and the project fails to meet expectations. This is especially true if the system was sold to policy makers on the basis of traffic improvements, which may be elusive, especially if traffic demand is increasing fast enough to mask any resulting improvements in signal timings.

Some agencies now require a business-plan approach for all their activities, and the systems engineering documentation will provide the necessary rigor to describe those activities, their objectives, and how those objectives will be measured.

Meaningful systems engineering is crucial to incorporating new technologies, as the above examples illustrate. An operations plan should provide an outline for a systems engineering process to be implemented for each routine and special project undertaken by the agency, following established approaches. These steps need not be excessively complicated. For example, many projects undertaken to support a particular activity might be described in a page or two. Even routine system projects can use a model concept of operations developed by the agency or within the industry, keying on the activities needed by that agency.

Sample Outline for a Traffic Signal Management Plan

Chapter 1. Objectives and Requirements.

In this chapter, define in concrete terms the operational objectives of the agency. While the definition should be concrete, it does not have to be specific. For example:

We will do our best to avoid making drivers stop, but when we must make them stop, we will delay them as little as possible, within the context of safe operation.

This mission statement is not specific, but it is concrete, and it can be used as the basis for establishing high-level requirements for the agency. Other objectives are possible, given local priorities, and based on the expectations of motorists in the region. The objective should avoid, however, specific performance standards, because traffic signals do not create capacity, and any given performance standard will fail when demand exceeds capacity.

Based on the objectives, the agency should establish high-level requirements that lead to strategies implemented by the agency. These strategies must be linked to the objective.These requirements may include:

Reliability standards, based on the percentage of time the motorists see what the agency intends for them to see.
Signal timing standards including:
- Minimizing congestion and managing unavoidable queuing.
- Promoting smooth flow
- Appropriate for conditions
- Versatile to provide acceptable operation across the range of conditions

Note that these requirements do not impose either a particular design or even a given level of resources. In the face of limited resources, creative design may fulfill these requirements, as described throughout this report. Some examples:

Improving reliability by improving design and initial construction quality, by minimizing the dependence on features and infrastructure (particularly detection) and by using system approaches that do not depend on fragile communications infrastructure.
Improving versatility by designing signal timings that are purposely broad-banded, such as progression-based solutions, rather than narrowly optimal as might be suggested by some optimization programs. Or, alternatively, strongly funded agencies might fulfill this requirement by implementing systems that continuously respond to changing traffic conditions based on measured inputs. Both of these fulfill this high-level requirement, but they do not both make the same resource demands.

This chapter does not need to include these strategies, but should establish the standard by which these strategies will be evaluated. These standards become the agency'acts definition of good, basic service.

Chapter 2. Responsiveness to Citizens, Media, Policy Makers, and Elected Officials

This chapter should lay out the standards defining the agency's responsiveness to motorists and other users of traffic signals. Examples of these standards include:

All citizen complaints will be responded to in kind. Telephone complaints will receive a telephone response, written complaints a written response, and so on.
All citizen complaints will be recorded by the person receiving the call, and responded to within a certain time period to be determined by the agency.
All complaints should be checked to rule out malfunctions first. This strategy provides a range of requirements for signal systems and their ability to provide fault and real-time monitoring.
Operational complaints that are not malfunctions should be addressed by the person who is responsible for the design. Explaining a design should never be delegated.
All requests requiring data collection and studies should be addressed within a given time period defined by the agency.
All complaints and requests should be tracked. The complexity of the tracking method is a strategy unique to each agency, but at the very least, the tracking method should ensure and be able to demonstrate that all complaints and requests are completely and expeditiously addressed. Most agencies already track requests coming from policy makers and elected officials, but the agency should also find ways to track common citizen complaints.
The agency should define how the traffic signal operations staff should or should not communicate with the media. Many agencies already have such policies in place, but within what is allowed by those policies, agencies should ensure that the highest priority is given to the story told to the public. That usually means providing the agency's most articulate, positive, and knowledgeable explainer to respond to media requests.
The agency should define standards for communicating program and project results to policy makers, elected officials, and the public. Those standards should include the format for such communications, an explanation of the objective of the program or project, the approach taken to achieve the objective, and the results in terms relevant to the objective. Agencies should describe here which programs and projects will be communicated this way, but it should at least include all programs and projects that receive specific funding, either within the budget or from outside sources.

Chapter 3. Maintenance Strategies to Achieve Objectives

This is where the agency lays out its maintenance approach. This chapter should include a clear explanation of how the agency will respond to limitations and constraints in maintenance resources, including limitations in staff numbers, training levels, and equipment. These strategies may include, for example, how an operational approach that does not depend on extensive detection minimizes the need for a large and difficult to maintain detection infrastructure. They should include how consistent design standards will be used to prevent sub-standard installations that require high maintenance levels. They should include how the use of features and technology within the signal equipment will be kept within the knowledge, skills, and abilities that can be expected for the job descriptions and remuneration for maintenance technicians.

Chapter 4. Operations Strategies to Achieve Objectives

In this chapter, the agency will describe the following:

The operational approach. Examples include maximizing progression, eliminating onset of yellow during platoon arrivals, pedestrian accommodation, clearance interval approach, basic detection strategy (consistent with the agency's resources), signalized left turn warrants, allowable phase sequence approaches, and so on.
Evaluation. How will the agency ensure that the operation on the street meets objectives, including operating according to design, providing smooth flow, providing sufficient versatility, and so on. The evaluation approaches can be quantitative (e.g. travel-time studies) or qualitative (field observation), but they must be directly linked to the objective. If the objective is to keep traffic moving, then the evaluation must assess how well traffic moves. If the objective is to maximize throughput through a bottleneck, then the evaluation should measure throughput. If the objective is to manage queuing at a bottleneck, then the evaluation should directly evaluate queuing. If the objective is to improve safety at a location, then the evaluation should include tracking crashes.

Chapter 5. Design Strategies to Achieve Objectives

Most agencies already have documents guiding infrastructure design requirements. But those documents do not usually set quality standards as they affect operations and maintenance, and this chapter provides an opportunity for agencies to do so. This section should apply consistent standards for traffic signal design, including street furniture, pedestrian accommodation, numbers and placement of signal heads, and so on. Not all agency operations staff are responsible for these design standards, in which case they should be written down here with an analysis of how the standard does or does not impose limitations on operations and maintenance. This provides the operations practitioner with an understanding of how design affects operations.

A most important section in this chapter concerns the implementation of automated systems, either for traffic signal operation or for management (including, for example, a citizen complaint tracking system). This chapter should lay out a meaningful systems engineering approach, including how each project (internal or external) will progress from a clear statement of needs through a demonstration of how the finished project supports those needs. The process should include at least the following steps:

The Concept of Operations, including how an agency will use the system. The agency should write this document internally, or be the primary audience for a document written by consultants. In all cases, it should focus on what they will do, not on what features they want.
Requirements. Each activity described in the Concept of Operations imposes one or more requirements. For example, an activity that includes an operator reviewing the real-time operation of the system leads to requirements that establish that real-time capability, and a display capability that shows the operator what they need to see to complete that activity. Those requirements become the standard for design.
Verification. In this step, the designer demonstrates how each element of the design fulfills the requirements. This term also describes how the system will be tested to prove that it conforms to the design.
Validation. Here the implementer demonstrates that the system indeed supports the activities described in the Concept of Operations.

previous | next