Model Systems Engineering Documents for Adaptive Signal Control Technology (ASCT) Systems
D. CONCEPT OF OPERATIONS GUIDANCE
LOGICAL WORKFLOW IN PREPARING THE CONCEPT OF OPERATIONS
While the layout of the Concept of Operations described in this guidance will provide a logical flow for the intended readers, it is generally not prepared in this sequence. As practical traffic engineers, it is generally preferable to describe at an early stage the operational scenarios envisioned by the system operators. After initially describing the limitations of the existing system, you should describe all the situations in which you expect the ASCT system to provide benefit, and how you expect the system to operate in each situation. After describing the operational scenarios, you will then be in a position to better describe the specific ASCT system and user needs, the alternative non-adaptive strategies considered and why they were discarded, and the envisioned adaptive system. Then you will be able to revise the operational scenarios so they are consistent with the statements of needs and provide clear examples of the expected operation.
The Concept of Operations will be organized in the following chapters, following the structure recommended in ANSI G-043-1992.
- Referenced documents
- User-oriented operational description
- Operational needs
- System overview
- Operational environment
- Support environment
- Operational scenarios
Once you have completed the Concept of Operation, use this checklist to confirm that all critical information has been included:
Is the reason for developing or procuring the system clearly stated?
Are all the stakeholders identified and their anticipated roles described? This should include anyone who will operate, maintain, build, manage, use, or otherwise be affected by the system.
Are alternative operational approaches (such as traffic responsive or time of day coordination) described and the selected approach justified?
Is the external environment described? Does it include required interfaces to existing systems?
Is the support environment described? Does it include maintenance?
Is the operational environment described?
Are there clear and complete descriptions of normal operational scenarios?
Are there clear and complete descriptions of maintenance and failure scenarios?
Do the scenarios include the viewpoints of all involved stakeholders? Do they make it clear who is doing what?
Are all constraints on the system identified?
The following sections describe how to develop each chapter of the Concept of Operation.
Chapter 1: Scope
1.1 Document Purpose and Scope
The first part of this chapter is a short statement of the purpose and scope of this document. This will briefly describe contents, intention, and audience. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table in Appendix B. These statements should be customized to explicitly apply to your situation. One or two paragraphs will normally suffice.
1.2 Project Purpose and Scope
The second part of this chapter gives a brief overview of the purpose and scope of the system to be built. It includes a high-level description; describes what area will be covered by the project; and identifies which agencies will be involved, either directly or through interfaces. Sample statements that may be used in this chapter are contained in the Concept of Operations Sample Statements table. These statements should be customized to explicitly describe your project. One or two paragraphs will usually suffice. This section should be written late in the process, after the envisioned system has been described. It will be a brief summary to introduce the reader to the proposed system.
The final section of this chapter will be a brief discussion of the proposed procurement process. The method of procurement should be determined early in this process, because it will have an impact on the format and content of the system requirements document.
Chapter 2: Referenced documents
This chapter is a place to list any supporting
documentation and other resources that are
useful in understanding the operations of
the system. This could include any documentation
of current operations and any strategic plans
that drive the goals of the system under development.
In particular, it should include documents
that define the overall goals and objectives
of your agency that will be supported by the
traffic signal system. This includes local
and regional transportation program and policy
documents and relevant inter-agency, management
and labor agreements and memoranda of understanding.
It should also include the regional and local
ITS Architecture and include relevant codes
and standards, such as ANSI, IEEE, NTCIP,
CFR and NEC. It should include references
to detailed documentation of any required
interfaces to other systems such as a regional
traffic conditions map or an Integrated Corridor
Management system. However, do not treat this
as a bibliography. Only include documents
that are referenced directly in the Concept
of Operations. Sample statements that may
be used in this appendix are contained in
the Concept of Operations samples table (Appendix
Chapter 3: User-Oriented Operational Description
This is a brief description aimed at non-technical readers who need an understanding of the current system or situation. It should say in only a few words what the existing system is, how it is currently used, what you are currently able to achieve with the system and (most importantly) what you want to do that can't currently be achieved with the system.
3.1 How Does the Existing System Work?
Describe the following aspects of the existing situation in words and figures.
Describe the nature of the existing road or network of traffic signals for which you want to consider adaptive operation. Is the capacity of the roads constant, or does it change during peak and off-peak times as a result of parking restrictions and/or reversible lanes? For example, is it an arterial road or a grid, are there several crossing arterials, are there freeway interchanges, is it an isolated intersection or a small group?
Describe the traffic conditions in the area. For example, is traffic highly directional or balanced; is it heavy only in commuter peaks or also at other times; are conditions relatively predictable or subject to unpredictable fluctuations, due to incidents and diversions; and are there major events that occur frequently at regular or irregular intervals? Include a brief description of pedestrian and public transit characteristics, and how they influence other traffic and the signal operation.
Is the variability limited to a single movement or phase? Does individual phase demand vary considerably from service interval to service interval, not only in when it occurs but also its amplitude?
Has there been an effort to accurately document and study these characteristics and map them to traditional solutions? The results and details of the study should be included.
What are the specific and documented traffic characteristics that are unsolvable with your current systems and approaches?
Describe in broad terms the likely grouping of the signals. Are the intersections sufficiently close that they may be coordinated together under some traffic conditions? Are there groups of intersections that are separated by a sufficiently large distance that they will never be coordinated together?
Describe the land use in the area. For example, is it residential, commercial, retail, industrial or a combination of these? Are there major, concentrated traffic generators with specific traffic patterns?
Describe the agencies involved in the operation of the traffic signal system. This should include the primary agency that operates the signals, other agencies whose signals are under the control of the system, other agencies whose signal systems operate in a coordinated manner with this system (but may not currently be connected to your system); and other agencies (not operating traffic signals) who are affected by the system (such as transit and fire departments) and have some control over the policy, procedures or operation of this system.
Existing Architecture and Infrastructure
Describe the existing system architecture. Provide an appropriate system network block diagram and describe the following elements, as applicable:
- TMC, location of servers, locations of workstations and any associated LAN or WAN
- Intermediate hubs and on-street masters, that are between servers and intersections
- Communications infrastructure, including media, bandwidth and protocols
- Detector locations and technology (e.g., loops, video and other technologies; stop line, advance and mid-block detection zones; number of lanes served by each detector; and any special purpose detectors, such as speed sensors and dual-channel loops with transponder detection capabilities)
- Functions of the existing system currently in use
- When the signals were last retimed and what were the results
Describe the existing Regional ITS Architecture. Show how the existing system fits into the existing Regional ITS Architecture. Include relevant block diagrams that show systems and agencies that are mentioned in subsequent chapters of the Concept of Operation. You may need to revise this section after the subsequent chapters have been completed and after the requirements have been completed.
3.2 What are the Limitations of the Existing
At this point, summarize the reasons why the existing operation is considered inadequate and therefore needs to be improved. This may include a brief description of operations or actions the operator would like to be able to implement in order to address various deficiencies or unsatisfactory operations, but cannot with the existing system. The reasons why these actions or operations cannot be implemented may also be mentioned.
3.3 How Should the System Be Improved?
Describe in broad terms the general approach to improving the system. Examples of actions or operations that may be desirable include: overcome jurisdictional boundaries that prevent signals from operating together; implement a system that can recognize changes in traffic patterns and react quickly; implement a system that more efficiently accommodates transit vehicles; implement a system that manages queues in critical locations; and implement a system that recognizes differing traffic conditions in various sections of the coordinated network and can make appropriate responses in each section.
At the end of this section, the reader should be able to clearly see the justification for the proposed changes.
Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
3.4 Statement of Objectives for the Improved
This section is focused on describing the operational objectives that will be satisfied by the envisioned adaptive operation. This should NOT describe the equipment but rather HOW the equipment will be used. To describe the operational objectives of the system, answer the following question.
What are the operational objectives for the signals to be coordinated?
- Smooth the flow of traffic along coordinated
- Maximize the throughput along coordinated routes
- Equitably serve adjacent land uses
- Manage queues, to prevent excessive queuing from
- Variable, with either a combination of these
objectives, or changing objectives at different
- For a critical isolated intersection, maximize
In answering this question you should not limit yourself to the current situation. Consider how the objective may change over time, as new development occurs in the area, the number of signals changes and the intensity of the traffic load changes. Consider the period of time over which these changes may occur and compare that with the expected life of the envisioned adaptive system. You may need to select more than one objective as being appropriate for the system.
This objective seeks to provide a green band along an arterial road, in one or both directions, with the relationship between the intersections arranged so that once a platoon starts moving it rarely slows or stops. This may involve holding a platoon at one intersection until it can be released and not experience downstream stops. It may also involve operating non-coordinated phases at a high degree of saturation (by using the shortest possible green), within a constraint of preventing or minimizing phase failures and overflow of turn bays with limited length, and with spare time in each cycle generally reverting to the coordinated phases.
This objective seeks to provide a broad green band along an arterial road, in one or both directions, to provide the maximum throughput along the coordinated route without causing unacceptable congestion or delay on the non-coordinated movements. The non-coordinated phases would typically be vehicle-actuated and operated at a high degree of saturation (by using the shortest possible green), within a constraint of preventing or minimizing phase failures and overflow of turn bays with limited length, and with spare time in each cycle generally reverting to the coordinated phases.
Traffic signals are often provided so that major traffic generators along a street can have safe and efficient access to and from the arterial. In these cases, the objective is to equitably serve all traffic movements at each intersection. At the same time, coordination is generally provided along the arterial, but not at the expense of accessibility to local land uses. An example is a suburban retail shopping district that generates significant demand for left-turn and side-street movements, with unpredictable demand characteristics during time periods that are not normally considered when developing traditional coordination plans.
Where there are closely spaced intersections, such as at a diamond interchange or within a tight grid network, and especially when a short block is fed by movements from various phases, the primary objective is to ensure that queues do not block upstream intersections or movements (such as occurs when a left turn bay spills over into adjacent lanes, or left turn queues exceed the intersection spacing at a tight diamond interchange). This often requires constraints on cycle lengths and phase lengths to ensure that a large platoon does not enter a short block if it must be stored within that block and wait for a subsequent green phase. It may also involve "gating" a movement, so that a movement is stored at an intersection simply to hold it in a location that has sufficient queuing capacity, even though other movements at the intersection may not require the green time. Multiple phase service may also be an effective tool in management of queues, especially for minor movements where queue overflows can cause problems for major movements.
It is often the case that different objectives are appropriate at different times of the day and under different traffic conditions. An arterial road that provides access between a freeway and large residential areas, but also has traffic generators such as retail centers and schools, may require an objective of providing a pipeline maximum throughput during the morning and evening peak periods, but provide access equity during business hours and on weekends, and minimizing stops during other off-peak times.
Maximize Isolated Intersection Efficiency
This objective applies to adaptive control of an isolated intersection. Intersection efficiency may be defined simply in terms of overall delay, or in a more complex objective function that considers stops and other operational parameters.
3.5 Description of Strategies To Be Applied
by the Improved System
This section describes the adaptive coordination and control strategies that may be employed to achieve the operational objectives.
Provide a Pipeline
Providing a pipeline along a coordinated route could support the two objectives of minimizing stops along a route and maximizing throughput along the route. The provision of a pipeline along a coordinated route can be achieved by a system based on a common cycle length, and also by a system that provides coordination bands toward and away from a critical intersection without using a common cycle length.
A traditional, cycle length based system would use detection to decide on the direction required for coordination (e.g., inbound, outbound, balanced), it would allow non-coordinated phases to be served and offsets to be set in a fashion that maximizes the width of the pipeline in the desired direction. It will allow phase sequence to be selected to allow the pipeline in both directions to be maximized (e.g., by using lead-lag phasing for left turn phases on the coordinated route).
If the network or arterial road has "resonant" cycle lengths, it will allow those cycle lengths to be identified and vary the cycle length to allow the measured demand to be accommodated within the associated pipeline. Resonant cycle lengths occur on streets with regular signal spacing, where a cycle length that is equal to or 2, 4 or 6 times the travel time between pairs of intersections is a feasible length. These cycle lengths can be described as "resonant" because the relationship between spacing, traffic speed, and cycle length allows platoons in both directions to be served by progression.
If the network or arterial does not have resonant cycle lengths, either because the resonant cycle is infeasible or because of uneven signal spacing, or if the traffic demand requires a higher cycle length to provide enough capacity in the pipeline, the system should allow the cycle length to be adjusted to provide a pipeline with sufficient capacity in one direction.
A non-cycle length based system will define the bandwidth of the pipeline to match the phase length of the coordinated phases at the critical intersection within a group. Then the phasing at other intersections will be timed so that green is provided on the coordinated route to accommodate the pipeline.
Distribute Phase Splits
To provide access equity, the demand for all phases
will be handled equitably by serving all movements
regularly and not providing preferential treatment
to coordinated movements to the extent that
delays and stops of other movements are significantly
increased. To do this a system would be optimizing
an objective function that seeks to balance
delays or some surrogate measure proportional
to delays. Strategies to prevent queue overflow
on minor movements may be needed. Typically,
a system that distributes green time may do
so on the basis of detector occupancy on competing
phases or by swapping time between phases
based on max-outs and gap-outs.
This strategy may also be applied to a critical isolated intersection in order to maximize intersection efficiency in terms of delays, stops or some composite objective function.
To manage queues, a system will allow the offsets between intersections to be set in a fashion that allows queues to be cleared at the end of each phase in blocks that are required to store queues during a subsequent phase. It will allow cycle length to be managed to limit queue sizes on designated movements. It will provide a means to control the locations where queues are allowed to form.
This strategy may also be applied to a critical isolated intersection that has limited queue storage on movements that, if they overflow, adversely affect movements in adjacent lanes and therefore reduce intersection efficiency.
A system that can accommodate several different coordination
strategies will allow each strategy to be
selected by appropriate measurements of traffic
conditions and will be able to be set up at
the operator's discretion to manage queues,
maximize throughput, minimize stops along
the maximize the coordination pipeline and
equitably serve all (or designated) movements.
3.6 Alternative Non-Adaptive Strategies Considered
This section starts with a list and description of the alternative, non-adaptive concepts examined. If you have an existing coordinated system, describe the features of the system that you are not currently using that could potentially be used to improve your operation. Explain why each of these features is not being used. If you have previously tried to implement improvements that have not been successful or are no longer employed for other reasons, describe them along with the reasons they were discontinued.
Many coordinated systems have the capability of using
traffic responsive pattern selection (TRPS)
to select and engage timing patterns based
on measured traffic conditions rather than
by a time of day (TOD) schedule. This should
be considered in this section.
Carefully consider the potential and limitations of TRPS in your situation. TRPS may be appropriate in situations where you are confident that the traffic conditions will never be outside the range of conditions for which you have prepared and stored timing patterns. The number of alternative plans must be sufficiently small that the pattern selection algorithms can discriminate between the traffic conditions applicable to each pattern. The system's measurement of traffic conditions must be able to include all movements that are affected by the different alternative patterns (or a subset of movements that are reliable indicators of all affected movements).
Complex Coordination Features
There are numerous features available in modern signal controllers and coordinated systems that are often not used. Following is a list of features that may be available within your existing system. You should examine each one and, if it is available, discuss whether or not it is applicable to your situation and whether it would provide the improvement you are seeking with an adaptive system. If you already use a feature within your existing operation and that feature is critical to the overall operation and must be retained, particularly when the adaptive system fails, that should be noted in this section. If not, these should be considered for employment as back up strategies in the event the adaptive system fails.
- Actuated coordination
- Use of vehicle actuation on non-coordinated
phases reduces unused or wasted green time
and broadens the green band, particularly
at intersections with low volume phases.
But this approach may cause platoons to
be released into the network earlier, resulting
in unexpected downstream stops.
- Multiple (repeat) phase service
- overflow of turning bays can be reduced
by operating a turning phase more than once
each cycle. This may also provide the opportunity
to better coordinate upstream or downstream
turning movements, and allow better coordination
in the non-peak direction by allowing more
flexibility in the placement of the through
phases in each direction on the coordinated
route. Another application of repeat service
is when traffic on a side street is light,
but the cycle length of the arterial is
constrained by the coordination objective,
it is sometimes possible to serve a single
vehicle and return to the coordinated phase,
then serve another later arrival on the
side street and again return to the coordinated
phase without adversely impacting the coordination
on the arterial. Features related to phase
re-service include conditional service or
more complex phasing/overlap arrangements.
Despite these features, multiple phase service
is often difficult and complicated to implement
using current signal controllers.
- Variable phase sequences
-Many agencies maximize the coordination
bands by using different phase sequences
(particularly leading or lagging left turns
on the coordinated route), during different
traffic conditions. Different phase sequences
may also be used on the side street phases
to manage queue lengths on the coordinated
route. If your agency has a policy that
prevents this operation, that should be
made clear in this section.
- Omit some phases in some plans
or at different cycle lengths -
When protected/permitted left turn phasing
is used, it is possible to omit the protected
phase under some circumstances, such as
at lower cycle lengths or in some coordination
plans. This is often used in conjunction
with a flashing yellow left turn arrow (FYLTA).
This technique provides a wider range of
possible cycle lengths in coordination patterns
and also allows more efficient free operation
at low volumes. If your agency has a policy
that prevents this operation, that should
be made clear in this section.
- Detector switching logic to change
the function of a detector - Detector
switching can improve intersection efficiency
by applying different logic when the controller
is in different states, such as to hold
the FYLTA on when the through movements
gap out early; to extend an overlap when
the demand for the overlap movement is greater
than the sum of the demands for the underlying
phases; or when different demand states
exist, such as calling or extending different
phases with and without pedestrian demands
- Coordinate different approaches
under different circumstances -
The appropriate phase to designate as the
coordinated phase is not always the through
phase in the peak direction. In some circumstances,
such as late-arriving platoons, it may be
more appropriate to designate the through
phase in the opposite direction.
- Coordinate turning movements
- In locations with an S-movement, particularly
when short block lengths or short turning
bays are involved, it may be most appropriate
to designate a turning phase as the coordinated
- Coordinate beginning or end of
green - The coordination strategy
you apply to the location should determine
whether the beginning or end of green is
used as the coordination reference point.
Minimizing stops often dictates use of the
beginning of green, while maximizing throughput
may dictate use of the end of green. In
particular, management of queues at closely
spaced intersections often requires use
of end of green. The selection of the appropriate
reference point should be made on an intersection-by-intersection
and pattern-by-pattern basis, rather than
as a blanket rule.
- Early release of hold
- It may be appropriate to allow the coordinated
phase to gap out early in order to better
serve the platoons in the opposite direction
during the next cycle, or better serve crossing
movements in a network with coordinated
cross streets. This may also improve coordination
in the peak direction at this intersection
and allow the bandwidth of the next cycle
to be wider.
- Hold the position of uncoordinated
phases within a cycle - In a network
with coordinated cross streets, it is often
desirable to hold the position of a cross
street phase in the cycle, rather than start
earlier if another phase is skipped or terminates
- Late phase introduction in coordination
- When traffic is light, there may be no
demand on a side street phase when it is
scheduled to run, and the phase is not introduced.
However, if a single vehicle arrives late,
rather than making it wait a complete cycle
for the next phase, it may be possible to
start the phase after its normal introduction
time and still return to the coordinated
phase at the correct time.
- Late pedestrian service
- This feature available on some controllers
allows the pedestrian walk to be introduced
after the phase is already green. This is
useful at intersections with high pedestrian
volumes, during times when the green time
is expected to be longer than the minimum
required for the pedestrian service.
- Stop-in-walk - This
allows for phase green times to be set lower
than the minimum time required for pedestrian
service. It takes some time away from the
start of the next phase but allows a lower
cycle length to be used. It is suitable
at locations at which pedestrians are not
present every cycle.
- Dynamic max - This feature
allows the splits to be changed if one or
more phases repeatedly run to maximum time
and does not gap out. This allows a non-coordinated
phase with a short peak that occurs when
there is spare time available on the coordinated
phases to be accommodated at a shorter cycle
length, by using a shorter maximum time
for that phase. However, if all phases are
running to their maximum time, the feature
has no effect. This may also be called critical
intersection control (CIC) or adaptive split
- Specified preemption operation
- This could include specific phase sequences
before and after the preemption event. It
may also include specific interval timings
as well as limited service during the event.
The call for preemption could be generated
locally or as directed from the central
signal systems or other means outside of
the traffic signal operational systems.
Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
Chapter 4: Operational Needs
This chapter lists the needs that will drive the requirements for the system. The system needs will be driven by the answers to the questions on the operational objectives and strategies, desired signal operational features, and the type of adaptive concept you plan to implement. The user needs will be driven by the answers to the questions about user interface, reporting and monitoring and maintenance requirements. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
Following is information that will guide your answers to the relevant questions.
4.1 Adaptive Strategies
Before starting to define the desired adaptive operation, it is useful to first review the various operational strategies and coordination techniques that will be considered in the following sections. The existing situation gives a good indication of what the most appropriate type of adaptive operation will be. So, based on the current signal control in the location(s) under consideration, or on the experience with previous attempts to provide coordination, answer the following question. Further explanation is provided below.
In the absence of adaptive control, what is the best operation?
- Coordination around a fixed cycle length
- Actuated, isolated operation
- Actuated operation with one or more intersections slaved from a critical intersection
The following descriptions represent three alternative ways of operating traffic signals. Any traffic signal operation will fall into one of these categories, although there will be many different situations within each category. If you consider that your situation cannot be defined within one of these categories, it is very unlikely that adaptive signal operation will be suitable.
Traditional Traffic Operational Strategies
Coordination around a Fixed Cycle Length
Fixed cycle length coordination patterns are typically selected by time of day schedules, by traffic responsive pattern selection, or a combination of both. The signal timing within the pattern may be fixed time (the phase splits are identical from cycle to cycle) or semi-actuated (some phases have their duration determined by detecting the presence of vehicles or pedestrians). Coordinated signals using a fixed cycle length cannot be fully actuated, because one or more phases are guaranteed to be green for some part of the cycle, regardless of the presence or absence of vehicles. However, in some systems it may be possible for all phases to be actuated, with part of the coordinated phase being flexible and subject to the presence of vehicles on that phase.
This form of coordination may apply to two or more signals. If your situation currently has coordinated signals, or the proposed situation will benefit from coordination, answer yes to this question. Also, if the spacing and feasible cycle lengths form a natural resonance that provides wide progression bands in both directions, answer yes to this question.
This situation may apply to one or more signals. In the case of a single, isolated signal, this is normally operated in free or vehicle-actuated mode, with detection on all phases. It relies on detection of vehicle gaps to terminate phases short of the pre-set phase maximum time. If you use alternative maximum times for different times of the day, then adaptive control of the intersection may improve the operation. If analysis of the intersection indicates that different cycle lengths and/or phase splits would be optimal for different times of the day, adaptive control of the intersection may improve the operation.
In the case of two or more signals that have natural cycle lengths that are markedly different, are sufficiently far apart that queues from one intersection do not affect the operation of another, and platoon dispersal is such that vehicles arrive at each intersection in relatively low density platoons, then isolated, vehicle-actuated control is most likely the most efficient operation. If all three of these conditions apply, then consider each intersection in isolation as described in the preceding paragraph. If not all three apply, then coordinated, actuated operation may provide a more efficient alternative.
There are many situations in which a critical intersection is close to one or more minor intersections and the key operating objective is to ensure that queuing from the minor intersections does not adversely impact the operation of the critical intersection. This is often the case when there are frontage roads or other streets close to a freeway interchange that is operated as a single intersection. In such cases, it is desirable to operate the minor intersections as slaves to the critical intersection. While traditionally this may be achieved using hard-wired holds and releases, this can become unwieldy or impossible if several signals are slaved from the one master controller. Some adaptive systems provide the ability to run the greens at minor intersections in a manner that ensures progression to and from the critical intersection even when it is operating in free, actuated mode.
Adaptive Coordination and Control Techniques
Common Cycle Length
There are several types of adaptive systems that use a common cycle length to coordinate groups of signals. Each of these may be implemented with vehicle actuation on some or all phases, or no vehicle actuation. An adaptive system may be able to implement one or more of the following approaches and the approach may be able to be selected on the basis of measured traffic conditions or a time-based scheduler.
- Use a cycle length that has been previously determined or implemented by another system, and optimize the phase splits and offsets within that cycle length. A previously determined cycle length may be appropriate when the network topology is such that there are one or more "resonant" cycle lengths that are multiples or sub-multiples of the travel time between important intersections.
- Calculate an appropriate cycle length based on traffic demand, and then calculate splits and offsets that are also suitable for the traffic demand. This may be used to select from several pre-determined cycle lengths that are suitable for the network topology.
- Calculate an appropriate green band width and a related set of offsets to accommodate traffic on a coordinated route, then select a cycle length and phase splits that are compatible with that green band.
Systems that use a common cycle length are referred to in these documents as "sequence-based," because when demand is present, each intersection displays the currently permitted phases in the same sequence from cycle to cycle.
Isolated Actuated Adaptive Operation
In this type of operation, an adaptive system may calculate an appropriate cycle length and splits for a signal, or calculate appropriate splits within a specified cycle length and apply them to an isolated signal that is not coordinated with any other signal. In the latter case, the cycle length may be specified according to a schedule. This operation may also be characterized as "sequence-based."
Actuated Linked Operation
This type of adaptive operation may or may not use a repetitive cycle length, and may allow a flexible sequence of phases. If the sequence of phases is flexible at all intersections, there is no cycle length, and it this operation is referred to in these documents as non-sequence-based. At the critical intersection, the sequence of phases and length of time provided for each phase will be determined by an objective function or other logic seeking to minimize delays and/or queue lengths in some fashion. This function may be definable by the operator. At the non-critical intersections, the green on coordinated routes will be displayed to provide progression of platoons traveling toward and/or away from the critical intersection. This may be done in a manner that optimizes some objective function such as number of arrivals during the green phase.
Unlike sequence-based and non-sequence-based operation described above, phase-based operation is generally restricted to a very small number of signals, typically one or two minor intersections close to one critical intersection. In this operation, phases at the minor intersections are generally timed in relation to specific phases at the critical intersection, while the critical intersection may operate in either vehicle-actuated mode or at a specific cycle length. This approach to coordination is often suitable for a freeway interchange with nearby frontage road intersections.
4.2 Network Characteristics
What is the size of the network that needs to operate adaptively, both initially and in the future?
If you are considering adaptive control for a group of signals, does this group include all the signals that you are likely to ever want to operate adaptively, or are you considering this a demonstration that, if successful, may be expanded to include other groups of signals? Answer the following questions:
- How many intersections in total need to operate in adaptive mode?
- Will the signals be divided into groups that will be expected to operate together in a coordinated fashion all or some of the time? If so, how many signals will be assigned in each group, and what will be the largest number of signals in one group?
- Will the number of signals in a group need to be flexible, or will it be constant?
- If the groups are widely dispersed, what is the distance between them? When the groups are dispersed, will they operate independently at all times?
- Describe the expected interactions and relationships between the groups.
Number of Signals
This is the total number of signals that may be operated under adaptive control, in all locations at which adaptive control may be considered. Depending on the size and responsibilities of your agency, this may include signals in different localities separated by hundreds of miles, or may simply be a small group of intersections on one arterial in a small community. The type of adaptive control you may have in mind or finally specify may differ from one location to another, and this will be considered in a subsequent step.
Size of Groups
Determine the number of intersections that may need to be controlled together as a logical group. Coordinated groups of signals may naturally be separated by distances that are sufficiently long that platoons tend to disperse and need to be regrouped to be handled effectively. The nature of a continuing arterial road may change to the extent that efficient operation in one section is quite different from efficient operation in another, such as changing from a four lane undivided road to a six lane divided road. Consideration of these factors will define the maximum number of signals that need to be controlled in one logical group under adaptive control.
In typical TOD coordinated systems, the grouping of signals often varies by time of day. For example, an arterial road may have the same cross-section for its entire length, but have several different and distinct sections, with different traffic characteristics. This may be illustrated by Ygnacio Valley Road, Walnut Creek, CA. At its western end, it connects to two freeways adjacent to the downtown and has several crossing arterials. Its central section has one crossing arterial and serves schools, a major medical center and small retail center. Its eastern section has one crossing arterial and serves a large business park adjacent to that intersection, with access to the business park via several signalized intersections on each of the two arterials. At different times during the peak periods, the heaviest movements may be westbound, eastbound, or relatively balanced, while the total volume is similar along its entire length. During business hours, the desired cycle length is different in each section. As a result, when the TOD signal timing plans are prepared, the signals may be grouped into one, two or three separate coordinated groups, depending on the desired cycle length, the volume of traffic in different sections, and the lengths of queues that appear at the boundaries of the groups if they are not synchronized.
Using the same logic, you may expect that the logical group you have defined may not be constant under all circumstances. Would you expect an adaptive system to vary the composition of the logical control groups under any of the following circumstances?
- If the system is cycle-based and the selected cycle lengths of adjacent groups are close, should they be forced to operate at the same cycle length?
- If queues at the boundaries between two groups may become long and interfere with the operation of the adjacent group, should they be forced to operate as one group?
- If the volume traveling between two groups exceeds some threshold, should they be forced to operate as one group?
Another example of the intersection grouping not being constant occurs when an arterial road or heavily used route lies within a grid network. During peak times, it may be common to operate the arterial road and the grid streets independently. However, during evenings and on weekends when traffic is lighter, the signal timing may be determined by the street geometry rather than the traffic volumes, and the most efficient operation may be to combine all intersections into one coordinated group.
Figure 8. Example of Flexible Grouping
Separation of Groups
If you have defined logical groups that are in widely separated locations, are there any circumstances under which they may need to be operated in a coordinated fashion? This may include synchronizing them under unusual circumstances, such as when there is major diversion of traffic due to an incident elsewhere, or during special events. At this time, do not consider whether they should be under the control of the same system; this will be considered in a later section.
Relationship of Groups to Other Groups or Facilities
Are there multiple groups of signals that impact (and interact with) each other? In some situations the groups may be parallel or perpendicular or crossing.
Will a group be impacted by major sources or sinks such as freeway interchanges, parking facilities, shopping malls, rail facilities, and event venues?
Describe the relationship of your agencies to those directly involved or immediately adjacent to the proposed adaptive system. Who are they and what are their expectations for the project and what can they bring to the table.
4.3 Institutional and System Boundaries
Often, the signals operated by one agency are adjacent to signals operated by another agency and they are timed to operate in a similar and compatible fashion. This may apply to groups of signals along one arterial road (e.g., crossing boundaries between adjacent cities), to groups of signals on crossing arterials (e.g., a city arterial crossing a state highway), to groups of signals in a grid network adjacent to an arterial road (e.g., a downtown area straddling a state highway). It may also apply to small groups of signals of one agency isolated within a larger group operated by another agency, such as interchange ramp signals on an arterial road. Answer the following question.
How will borders with other systems and institutions be handled?
- Make neighboring system part of the operation, but remain separate system
- Make neighboring system part of the system
- Make the adaptive system respond to the operation of the neighboring system
- Allow stratified levels of authority and control
These situations are very similar to the considerations of the relationship between an adaptive system and crossing arterials.
Part of the Operation
In this arrangement, the signals operating in the adjacent system would not be part of the adaptive system. However, they would operate in a manner consistent with the adaptive system, so that coordination is not broken at the boundary between the two systems. This could be achieved by providing communication between the two systems, and have the adaptive system transmit key operational information such as cycle length, time clock synchronization data and offset data, so the non-adaptive system could adjust its operation in response to changes in the adaptive system's operation.
Part of the System
In this arrangement, the signals operating in the adjacent system would become part of the adaptive system.
Constrained adaptive operation
In this arrangement, the adaptive signals would operate within constraints imposed by the adjacent system, so that coordination is not broken at the boundary between the two systems. For example, the adaptive system may operate on the same cycle length that is in operation in the adjacent system, and optimize its phase splits and offsets along its coordinated route. This could be achieved by receiving data from the other system(s) on cycle length, synch points and offset requirements. It may also be achieved by sensing the traffic conditions within the other system(s) and adjusting the adaptive operation to accommodate that traffic.
Crossing Arterial Coordination
Often an arterial road that is coordinated will be also coordinated with other arterials and streets that cross it or it may be part of a grid network whose signal timing is determined by that of the arterial. If the route for which you are considering arterial operation is currently also coordinated with cross streets or an adjacent grid, you need to consider whether that relationship should continue with the adaptive system. Answer the following question.
If there are crossing arterials, how will compatibility be maintained?
- Make them part of the operation, but remain separate systems
- Make them part of the system
- Constrain the adaptive system to respond to the operation of the crossing arterial
Part of the Operation
In this arrangement, the signals operating
on the crossing arterials would not be part
of the adaptive system. However, they would
operate in a manner consistent with the adaptive
system. This could be achieved by providing
communication between the two systems and
having the adaptive system transmit key operational
information such as cycle length, synch point
and offset data, so the non-adaptive system
could adjust its operation in response to
changes in the adaptive system's operation.
Part of the System
In this arrangement, the signals operating on the crossing arterials would be included within the adaptive system.
Constrained Adaptive Operation
In this arrangement, the adaptive signals would operate within constraints imposed by the systems controlling the crossing arterials. For example, the adaptive system may operate on the same cycle length that is in operation on the crossing arterials, and optimize its phase splits and offsets along its coordinated route. This could be achieved by receiving data from the other system(s) on cycle length, synch point and offset requirements. It may also be achieved by sensing the traffic conditions within the other system(s) and adjusting the adaptive operation to accommodate that traffic.
Adaptive Coordination Strategies
Part of the Operation
The adaptive system optimizes its operation on an arterial in accordance with its own objective functions. It then communicates key data to the adjacent systems operating other intersections, and those systems adjust their operation to maintain consistency with the adaptive system. This would typically involve sending cycle length, synch point and offset data to the other systems.
Part of the System
In this arrangement, there would be one integrated adaptive system controlling all the arterials.
Constrained Adaptive Operation
There are two ways in which an adaptive system could operate or be constrained to operate in concert with an adjacent system over which the adaptive system does not exert influence.
- Use adaptive control that does not vary cycle
length, but accepts cycle length as an input. The
adaptive system will then adjust splits and offsets
at the intersections along its coordinated route,
based on the measured traffic conditions.
- Detect platoons leaving the adjacent system and
approaching the adaptive system. Adjust the adaptive
system's operation to accommodate the arrival of
There will be staff in different sections of your agency, and in different agencies, who will need to have access to the system to perform different functions. For example, an operator will need to be able to modify parameters to be able to set up and fine tune the operation of the system. Other operations staff will need to be able to generate reports of system performance, but not need to modify parameters. Operations staff in other agencies may need to monitor the traffic conditions within the adaptive system, and could be given permission to modify its operation under certain circumstances. To achieve this arrangement, the system needs to have assignable security levels and jurisdiction rights that allow an administrator to assign appropriate privileges to each potential user.
4.5 Queuing interactions
The presence of queues within the area proposed for
adaptive control, or the risk of adaptive
operation causing queues that affect other
elements of the transportation system, are
very important to the required capabilities
of the envisioned system. Answer the following
Are some queues outside the control of the envisioned system?
- Will queues back into the system from downstream congestion?
- Do queues form within the system from traffic generators rather than due to signal operation?
- Will queues that propagate outside the system be unacceptable?
- Is it important to control where queues are stored within the system boundaries?
- Is there a need to flush queues through the system?
Will queues back into the system from downstream congestion?
A common situation on an arterial road that is serving
freeway on-ramps is for congestion on the
freeway or queuing from a ramp meter signal
to cause a queue that may block an intersection
within the adaptive group. In these cases
it is necessary to detect the presence of
a queue before it causes an adverse impact
on the adaptive system and adjust the system's
operation to prevent the queue extending into
Do queues form within the system from traffic generators rather than due to signal operation?
If a queue builds up at a location within the network that cannot be controlled by the adaptive system, such as at the entrance to a parking garage during a major event, it is generally desirable to detect the queue and modify the system's operation to prevent the queue growing to an extent that it compromises the system's operation.
Will queues that propagate outside the system be unacceptable?
At locations such as an intersection with a freeway off-ramp, excessive queuing on the off-ramp may have consequences that outweigh the objective of most efficiently coordinating an arterial. In such cases, it would be desirable to detect the presence of queuing at entry points to the adaptive system and modify its operation to prevent the queue from growing.
Is it important to control where queues are stored within the system boundaries?
If there are short blocks between adjacent signals, such as at freeway interchanges, it is often necessary to ensure that some approaches are always clear of queues at the end of each phase. To do this often requires the green time at one intersection to be controlled so that queues form at that intersection and are only released downstream when they can clear the next intersection. This requires the adaptive system to have the ability to control the green time and the offsets so queues are stored at the desired intersection and released at the desired time. Queues may also be prevented from overflowing by use of multiple phase service. If preventing queue overflows from specific movements requires more frequent service than is feasible for all movements, then multiple phase service is needed.
Is there a need to flush queues through the system?
This consideration is related to the previous, in that the system needs to be able to ensure that once a queue is released it is not stopped until it passes a designated intersection.
How should pedestrians be accommodated by the adaptive system?
- Are pedestrians central to the operation?
- Special features (e.g., pedestrian early start, variable lane/movement assignment, variable turn restrictions, programmed pedestrian recall)
- What adaptive response is required?
- What current techniques must be retained?
This question will help you decide to what extent the accommodation of pedestrians by the system will impact your selection of an adaptive system. If your system has little expectation of pedestrian activity, then accommodating or optimizing the operation with the pedestrians intervals included is not needed. Rare pedestrian activity can be accommodated by the local controller even if it must override adaptive operation in these cases, with no overall detrimental effect. But if pedestrian activity is frequent during critical periods, the adaptive control should accommodate the pedestrians and optimize the operation around them. If pedestrians are accommodated using more complex features, you will need to carefully define the manner in which pedestrians must be accommodated by the adaptive system.
Examples of more complex features include: changing the phasing and/or allowable movements on an approach to a signalized intersection when the presence of pedestrians causes unacceptable queuing; and coordinating phases that serve pedestrians at adjacent closely spaced intersections.
Do you need the adaptive system to:
- Disregard pedestrians when calculating the adaptive signal timing, but serve rare pedestrians locally, ignoring the adaptive timing
- Incorporate accommodation for occasional pedestrians in the adaptive signal timing
- Always allow the full pedestrian time in each phase
- Allow custom pedestrian features
- Some or all of the above, depending on the situation.
In this section, give a brief overview of pedestrian operational needs. The details will be examined as the requirements are defined. The Concept of Operations table contains sample statements that may be used.
4.7 Non-Adaptive Situations
From time to time it is possible that traffic conditions would fall outside the range of conditions that an adaptive system can accommodate. When these conditions occur, one of the fallback non-adaptive modes of operation should be employed. There are various ways in which this may occur.
Are there situations that may lead you to sometimes require non-adaptive control? Do you want the system to:
- Deal with it adaptively and automatically?
- Go to non-adaptive control during the presence of a defined condition (such as exceeding volume and occupancy criteria on specified detectors)?
- Operate non-adaptively according to a user-defined schedule?
- Operate non-adaptively during special events and diversion around incidents?
Adaptively and Automatically
In this situation, the adaptive system would be capable of adopting a wide range of operation covering the full range of expected traffic conditions. The system's objectives would seamlessly move from one to another depending on the measure traffic conditions, and its operation would change to suit those conditions without manual intervention.
In this situation, the adaptive system would recognize that the conditions are outside its operating range and automatically choose a non-adaptive mode of operation while the conditions remain in this state. It would automatically revert to adaptive operation when conditions returned to the acceptable range. An example of this situation is diversion along a corridor arterial around a freeway incident. The adaptive system would detect that the incident has occurred, or detect the diverting traffic itself.
This situation would involve using a time of day scheduler to direct the system to operate in a non-adaptive mode. This may be appropriate when there are predictable conditions to which the adaptive system may not be able to react, or may react more slowly than necessary. An example of such a situation is the change of shift at a large factory, or end of classes at a large school, both of which often result in a sudden and predictable increase in traffic volume exiting a parking lot.
In this situation, an operator would manually force the system operation to a non-adaptive mode. Typically, this would require operators to use stored plans that had been previously developed, or manually modify operation in real time in response to observed conditions.
4.8 System Responsiveness
Depending on the nature of the traffic within the system's area, you may or may not want the adaptive system to continually adjust to small changes in the traffic demands. Answer the following question.
How responsive do you want the system to be to:
- Small shifts in demand?
- Large shifts in demand?
Small Shifts in Demand
Depending on the nature of the traffic within the system's area, you may or may not want the adaptive system to continually adjust to small changes in the traffic demands. To some extent this reflects the different capabilities of the various systems, but also will depend on the type of road network that will be coordinated. For example, a system that maintains a constant cycle length can efficiently vary splits from cycle to cycle. A system that constantly adjusts cycle length can also efficiently react to small changes in demand. However, if the arterial road or network has one or more "resonant" cycle lengths, small changes in demand will not require changes in the underlying cycle length.
Large Shifts in Demand
Large shifts in demand should be detected and accommodated by all adaptive systems. However, the nature of the traffic conditions will determine whether the system will need to react quickly to sudden large changes, or gradually. If large shifts in demand occur relatively gradually over a period of time, such as the buildup and decay of most peak periods, then that will define the rate at which the adaptive system should be expected to react. However, if the system operates in an environment in which the level of demand can vary dramatically without a predictable schedule, such as emptying a parking lot at the end of a large sporting event or accommodating traffic diverted from a freeway as the result of a peak hour incident, then you may require the system to detect the change and react swiftly. If the change in demand level is likely to be observed by an operator, then this rapid response may not be required.
It may also be appropriate to have different response times required for adaptive solutions and non-adaptive solutions to measured conditions. For example, if a large shift in demand is detected, and the condition is outside the range of acceptable conditions, it may be desirable to immediately implement the non-adaptive solution. However, within the acceptable range, a slower response may be more acceptable in order to prevent oscillations or instability in the system's operation.
You may wish to define different rates of response to different conditions. Several examples are included in the Concept of Operations sample statements and include such situations as: the normal rise and fall of volume around peak periods; rapid increases in volume during diversion from an incident, change of shift at a factory, or end of the school day at a high school; the detection of spillback of a left turn queue or a ramp metering signal; the prediction of queue spillback through recognition of repeated phase failures.
4.9 Complex Coordination and Controller Features
Answer each of the following questions to determine what advanced controller functions need to be maintained during adaptive operation. Each of these items is described in more detail below.
- Do you need multiple phase service?
- Do you have multiple overlap phases?
- Do you permit variable phase sequences?
- Do you omit some phases in some plans or at different cycle lengths?
- Do you use detector switching logic to change the function of a detector?
- Do you have non-standard features? If so, identify them.
- Do you coordinate different approaches under different circumstances?
- Do you coordinate turning movements?
- Do you require early release of hold?
- Do you require the ability to hold the position of uncoordinated phases within a cycle?
- Do you allow late phase introduction in coordination?
- Do you use protected/permissive phasing?
- Do you require Flashing Yellow Arrow protected/permissive and permissive only left turn service?
- Do you require movement restrictions
by volume or time-of-day? For example a
left turn restriction (possibly phase omit)
- Do you have special requirements for one lane of an approach?
- Do you require transit queue jump operation by TOD or adaptive?
- Do you require specific phases and sequences to occur following a pre-emption event?
Multiple Phase Service
Multiple phase service may be required for a left turn phase that has a short turning bay that would overflow if the turning phase was simply leading or lagging the opposing through movement. In this case it is efficient to run the left turn phase both before and after the opposing through phase. This may operate with both protected only left turns and protected/permissive left turns.
When traffic on a side street is light, but the cycle length of the arterial is constrained by the coordination objective, it is sometimes possible to serve a single vehicle and return to the coordinated phase, then serve another later arrival on the side street and again return to the coordinated phase without adversely impacting the coordination on the arterial.
This section applies to situations where the multiple phase service is provided by the signal controller and must be tolerated and accommodated by the adaptive system.
Do you have multiple overlap phases?
An intersection with complex channelization may have multiple overlap phases to accommodate heavy turning movements, rather than simply relying on right turn on red to minimize delays to right turning traffic.
Do you permit variable phase sequences?
Many agencies maximize the coordination bands by using different phase sequences (particularly leading or lagging left turns on the coordinated route), during different traffic conditions. Different phase sequences may also be used on the side street phases to manage queue lengths on the coordinated route. This would also include actuated extended pedestrian phase intervals.
Do you omit some phases in some plans or at different cycle lengths?
Do you use protected/permitted left turn phasing, and omit (exclude) the protected phase under some circumstances, such as at lower cycle lengths or in some coordination plans?
Do you use detector switching logic to change the function of a detector?
Do you use logic to change the phase call and extend functions of a detector, based on the state of the signals and the state of demands for other phases? If so, describe the logic.
Do you have non-standard features?
Are there special operating features that you need to retain during adaptive operation that have been customized for your situation? For example, an intersection with a wide central median may have pedestrians cross in two stages, or may operate the pedestrian crosswalks on each side of the median independently, and may overlap with vehicle phases that would normally be in conflict in an 8-phase operation. Describe any such features.
Do you coordinate different approaches under different circumstances?
Do you sometimes coordinate the through movements on one road, and at other times coordinate the movements on the crossing street at the same intersection?
Do you coordinate turning movements?
Do you sometimes coordinate a turning movement rather than a through movement on an arterial road? There are several relevant examples of this situation. If you have the coordinated route with the heavy movements turning left or right at an intersection, it may be desirable for the left turn phase at that intersection to be the coordinated phase. The coordinated route through a network or in one section of an arterial may involve a heavy S-movement, which may require the coordinated phases at two adjacent intersections to be turning phases, or one turning phase and one overlap.
Do you allocate unused green time to a non-coordinated phase?
At a critical intersection on an arterial road, or at the crossing point of two arterials or two coordinated routes in a network, it is sometimes desirable to allow unused green time from turning phases to be added to through movements that are not the coordinated phase.
Do you need the ability to choose which phase to allocate unused time to? For example, you may wish to add unused time to the next phase, to a later uncoordinated phase, or to the beginning of the next coordinated phase.
Do you have special requirements for one lane of an approach?
Do you have downstream movements that result in queuing in one lane of an approach while the adjacent lanes are not congested? A typical situation of this nature occurs when there is a heavy right turn onto a freeway ramp. The queue for the right turn may extend upstream through several intersections in one lane only. In such a situation, there may be no benefit in extending the green for that movement, so the adaptive system would need to be able to discriminate between lanes and react differently to demand in each lane.
Early Release of Hold
It is sometimes desirable to allow a coordinated phase to terminate early, once a platoon has passed through the intersection, so that unused green time can be added to the beginning of the next coordinated band. This often improves the coordination in the non-peak direction.
Hold the Position of Uncoordinated Phases
At times it is desirable that if a phase has no demand, the preceding phase continues green rather than immediately skipping to the next phase. This is known by various terms such as, sliding yield window, yield by phase, multiband permissive and false green.
Late Phase Introduction
When the cycle length is determined by the network geometry and the traffic volume on side streets is light, the "pipeline" along the arterial route is often not fully utilized and the side streets often gap out and return to the coordinated phase early. When there is no pedestrian call on the side street and no vehicle call at the yield point of the coordinated phase, the side street phase is typically skipped. However, some systems allow the side street phase to be introduced later in the cycle if there is still no pedestrian call and a late-arriving vehicle could be served without forcing the intersection out of coordination.
4.10 Monitoring and Control
What form of monitoring and system control is required?
- From a central TMC?
- On-site (at controller or on-street master)?
- From multiple TMCs?
- From remote locations (not from remote TMC)?
- From maintenance vehicles?
- By an Integrated Corridor Mobility or other external system?
This refers to the location at which an operator will be able to observe and control the system's operation, using a workstation, laptop, smartphone, etc. It does not refer to the physical location at which the system's equipment is located. The alternatives typically required by an agency include: from a workstation at a TMC; via laptop connected directly to a local signal controller or an on-street master; or from another remote location, such as elsewhere within the agency (such as the signal shop or traffic engineer's desk) via a WAN, or via the internet.
Describe this need in broad terms. The details will be examined as the requirements are defined. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
4.11 Performance Reporting
Your existing system may have the capability to provide numerous reports about its performance and day to day operation. These reports support management of the system, assessment of its performance, trouble shooting, and legal records. An adaptive system may replace the existing system, be a stand-alone addition or be integrated with your existing system; this may affect the reports and records that are available to you.
Do you want...
- To report measures of performance against the system's objective functions?
- To report measures of performance against your agency's mobility objectives?
- Real time logging of system status?
- Recording of events and data generated by the system (if so, what retention policies)?
- Data storage and data analysis, within or external to the system?
There are two parts to performance measurement:
- Is the system optimizing performance in terms of
its objective function?
- Is the effect of the system on traffic flows meeting
your mobility objectives?
The adaptive system will make calculations, and take actions based on the results of those calculations. To what extent do you wish to see the results of those calculations, which may be useful in calibration and fine tuning? For example, a system that calculates a requested cycle length for each intersection in a group, based on measured occupancy may report the occupancy of each loop, the requested cycle length at each intersection, the selected cycle length for the whole group, and the constraints that were applied in selecting that cycle length.
A system that adjusts phase times based on whether a phase gaps out or runs to its maximum may report the actual green time, the permitted maximum green time, the next maximum green time, and any constrain that influenced the calculation of the next maximum green time.
How are your agency-wide mobility objectives defined? If mobility is defined in terms of travel times, level of service and throughput, do you expect the system to report traffic conditions in those terms? If the objective function is to minimize delay or stops, are delay or stops reported by the system?
Other parameters that may be useful in reporting performance include: arrivals on green and red; green time utilization; and measured and estimated queue lengths. While these are useful, you should carefully describe your need in terms of the traffic or mobility performance you must report. However, you should also define the need in terms of the signal timing objectives you set for the system, such as minimizing phase failures, minimizing stops along a specified route, or controlling maximum delay to different users.
Real time logging involves recording such items as phase times, plan in effect, transitions, detections, cycle length, offsets and all alarms and events at the time they occur. It may also log the results of calculations and parameters that control the adaptive systems operation. In this way it is possible to see in real time, on a second by second or cycle by cycle basis the status of each signal. To what extent do you need this data available as it occurs or is collected, at the end of the current cycle (if applicable) or within a specified time interval?
What operational data do you need recorded and stored, and for what period of time must it be available within the system? Do you have an external data storage and analysis facility with which the adaptive system must interface and transfer data? What data is required by that system, and at what frequency must this data be transferred?
Will the performance measures be passed to another system, such as an Integrated Corridor Management decision support system, or a real-time regional traffic conditions map? Describe this need in broad terms. The details will be examined as the requirements are defined. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
An important guide for this issue is the regional ITS architecture. Show how any external interfaces are consistent with the regional ITS architecture, and identify any changes to that architecture that will be required a result of stakeholder agreements made during the preparation of this Concept of Operation.
Use Historical Gata to Recreate Events
Use historical logged data to study and investigate past situations. This would be used to address complaints such as "I sat at this approach for 30 minutes waiting for a green light" or "I had to stop at a red light at every intersection between here and there." This information is also important to determining if the system is responding to situations and input data as according to design expectations. Data warehousing is typically included in the regional ITS architecture. In this section, describe how the adaptive system will need to integrate with any relevant data warehouse.
4.12 Failure Notification
Your existing system has the capability to detect failures and report them in real time to appropriate staff for attention. Some adaptive systems require the operator to be logged in to the system to retrieve alarms and alerts, while others push this information out in real time.
How will notification of failure of the adaptive system be managed?
- Report the alarms and alerts directly to the operations and maintenance staff
- Interface with another system, which will in turn
report alarms and alerts through its notification
In this arrangement, the adaptive system would have its own system to notify maintenance and operations staff of alarms and alerts that it generates.
Report to Interfaced Dystem
In this arrangement, the adaptive system would send its alarms and alerts to another system that has a procedure for notification of maintenance and operations staff. This may be a parallel traffic management system, or a separate maintenance management system.
Describe this need in broad terms. The details will be examined as the requirements are defined. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
4.13 Preemption and Priority
In this section, you define the types of preemption and lower level priority that exist and need to be maintained, or will be required in the future.
Is preemption or transit priority required to operate during adaptive operation?
What forms of signal preemption and transit priority will be required?
- Railroad preemption
- Emergency vehicle preemption
- Light Rail Transit priority
- Bus priority
How will railroad preemption be accommodated by the adaptive system? Does the adaptive system acknowledge the preemption and have a suitable recovery process? Describe the preemption operation you need during adaptive operation. Include any special rules you use, such as which phase to serve during and after the preemption, and any additional logic based on measured queuing or other traffic conditions.
Emergency Vehicle Preemption
Does the adaptive system acknowledge and permit emergency vehicle preemption? Does it have a suitable recovery process? Describe the EV preemption operation you need during adaptive operation. Include any special rules you use, such as which phase to serve during and after the preemption, and any additional logic based on measured queuing or other traffic conditions.
There are several different methods of providing transit priority for buses and light rail. Some are entirely locally based, with communication directly between the transit vehicle and the signal controller, and all priority logic is resident in the local controller.
Other methods of implementing bus priority involve a separate system that determines the response to a priority request and communicates with the local controller. Will the adaptive system be required to communicate with an external priority system in order to provide transit priority? Describe the transit priority operation you need during adaptive operation.
Many LRT systems include logic and hardware to coordinate the interaction between the LRT vehicles and the traffic signal indications. Many of these systems also have special dedicated LRT phases within the traffic signal sequences. Some actually drive the traffic signal operations. Describe the transit priority operation you need during adaptive operation. Include any special rules you use, such as which phase to serve during and after the priority, and any additional logic based on measured queues or other traffic conditions.
Preemption / Priority Frequency
Document the number of preemption/priority calls that will occur under varying circumstances. Document any specific implementation rules that will need to be accommodated by the adaptive system.
Describe this need in broad terms. The details will be examined as the requirements are defined. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
4.14 Failure and Fallback Modes
This section will describe the behavior of the adaptive system and the non-adaptive components of the system in the event of failures of system elements that are important to efficient and reliable adaptive operation. Answer the following question.
What failure modes are required?
In the event that the adaptive operation cannot continue,
should the operation revert to an existing
TOD coordinated operation or free operation
of local signals? If TOD coordination is required,
should this be under control of a central
system, an on-street master or local intersection
controller database? Mention the extent to
which detector failures need to be accommodated
based on your agency's ability to maintain
detectors in an operable state.
Describe this need in broad terms. The details will be examined as the requirements are defined. Sample statements that may be used in this chapter are contained in the Concept of Operations samples table (Appendix B).
4.15 Definition and Application of Constraints
Now is the time to identify constraints that may limit the choices that are open to you. These constraints will either be accepted or you can choose to overcome the constraint. This will involve determining what the financial, resource or political cost to overcome the constraint will be. You will then be in a position to evaluate the trade-off between the benefits that will not be realized if the constraint is accepted and the cost of removing the constraint to achieve those benefits.
The following sections describe constraints that may be applicable to your situation.
Your organization may have existing policies in place that affect the equipment that may be purchased. There may be constraints placed on your organization by funding agencies. These should be identified, the implications for system selection quantified and the consequences of varying from those policies documented. Examples of infrastructure policies that may be applicable include:
- Controller type
- Detector type
- Signal system
- Communications media and protocols
- Cabinet type/space
Management and Human Resources
There are often explicit and implicit management and human resource policies that may become constraints. Great care should be taken with this category of constraint. There is often resistance to change that stems from lack of understanding and the "fear of the unknown." There are also many misconceptions about the nature of adaptive systems upon which staff at all levels form opinions that are difficult to change. Examples of management and HR policies that may be applicable include:
- Staff fears that their jobs will be reduced or made obsolete by the adaptive system
- Staff fears of additional workload without reward
- Agency hiring policies
- Agency training policies
Your available financial resources may have an impact on the needs that you are able to accommodate within your requirements. If you are preparing this Concept of Operations at the planning stage, in order to develop and program a project, then your financial constraint may be the maximum funding you expect is possible. This may be spread over several program years, which will give you guidance on when you could schedule the project, or it may give you guidance on how the project could be implemented in phases.
If you are preparing this Concept of Operations after a project has been programmed and budgeted, then this will give you guidance on the how complex the system could be, the geographical area that may be covered by the project, the extent to which the support environment can be implemented. If capital funds are limited, careful consideration should be given to the impact these decisions (limiting work to limit initial capital cost) will have on both short and long term operations and maintenance costs and efficiency.
With what level of complexity are you comfortable? You will need to consider the level of complexity of the various adaptive systems and relate them to the ability of your organization to handle that complexity. While it is difficult to quantify this factor, a useful benchmark for comparison that will be understood by most organizations that are considering adaptive operation is the level of complexity of conventional time of day coordination systems. Is the level of complexity your organization will be comfortable with:
- Similar to conventional TOD, without traffic responsive pattern selection
- Similar to conventional TOD, with traffic responsive pattern selection
- Somewhat more complex than conventional TOD
- Significantly more complex than conventional TOD
While this does not lead directly to system requirements, it does provide a basis for determining whether the skills and expertise of your existing operations and maintenance staff will be adequate for the selected system. This will lead to further decisions about appropriate qualifications of staff and training requirements.
This section will help you quantify the options available for operation and maintenance of the adaptive system within your organization.
- What are the capabilities of your staff (e.g., hardware maintenance in the field (including detection and communication), and in the office; signal timing skills)?
- What is your operations and maintenance structure (e.g., are signal technicians specialists or do they have other electrician duties; are signal maintenance staff under the direction of traffic engineering staff, or separate public works organization; are maintenance staff in-house or contracted?)?
- How is maintenance funded? Does it have a self-sustaining structure or does some maintenance need to be built into purchase price and warranty?
While this does not lead directly to system requirements, it does provide a basis for determining whether other organizational, administrative and financial changes will be required to accommodate the adaptive operation.
Hardware and Software Constraints
This section will help you understand the tradeoffs between flexibility to choose from all available products and limiting that flexibility to use a particular hardware or software product along with the adaptive system. In some cases, state agencies require the acquisition of traffic signal controllers based on a statewide selection. If you are a local agency and are subject to this requirement, then this constraint may be unavoidable.
If the requirement imposed by a state agency is in the form of an incentive, however, then the value of the incentive should be weighed against the consequences as revealed by this constraint preventing you from answering the questions in this document as you would wish based on your operational objectives.
Some agencies, however, might consider the acquisition of products that work with their favorite signal control equipment, or their current equipment, to avoid having to replace it or to avoid the complexity of using different signal controllers. If this applies to you, then consider whether this constraint is preventing you from answering the questions in this document as you would wish based on your operational objectives. If the constraint can be relieved, you may enjoy the benefits of a wider more competitive selection and of selecting requirements that are more closely linked to your objectives.
While you may have made commitments to others about the date by which your system will be operational, this should not be imposed as a hard constraint that could potentially compromise your selection of the most appropriate system. However, examples of schedule constraints that would be appropriate to impose on your procurement are:
- A major event for which the proposed system is
required to efficiently control traffic
- A hard date by which funding would be withdrawn
if sufficient progress is not demonstrated
Although it is appropriate to document these dates in the Concept of Operation, they should not be directly reflected in the System Requirements.
Chapter 5: Envisioned Adaptive System Overview
This chapter is an overview of the envisioned
adaptive system. It is a high-level description
that will describe the main features and capabilities,
other systems with which it will be interfaced,
and the scope of its coverage. You should
describe its conceptual architecture at a
block diagram level with a high-level data
flow diagram. This should not show design
This description should reflect the needs that are described in the previous chapter. It should illustrate, either graphically or in words, each of the following categories of needs that are relevant:
- Network characteristics
- Type of adaptive operation
- Interfaces to other systems
A good way of illustrating the system is to draw out the activities undertaken by stakeholders in a particular situation, and highlight those that are anticipated to be automated with the adaptive operation. An example of such a diagram is illustrated in Figure 9.
Figure 9. Sample System Block Diagram
Chapter 6: Adaptive Operational Environment
This chapter describes both the operational environment and the physical environment within which the adaptive system will operate.
6.1 Operational Environment
Describe the stakeholders. These should include all existing stakeholders who have an influence on the operation of the existing signal system. This will include traffic engineers involved in signal timing, TMC operations staff, and staff of other agencies whose operation and duties may be affected by the envisioned adaptive system.
The activities related to adaptive operation should be described, such as preparation of timing parameters, implementation and fine tuning, system performance monitoring, and inter-agency staff interactions.
The organizational structure should be described, highlighting any changes from the existing arrangements that are envisioned. An overview of the qualifications and experience of personnel should be presented, along with clear definition of any roles and responsibilities that would be undertaken by contractors, vendors, consultants and staff of other agencies.
Sample statements that may be used in this section
are contained in the Concept of Operations samples
6.2 Physical Environment
You should describe the facilities within which equipment and personnel will be housed, additional furniture and equipment that will be required, new computing hardware and software that will be required, operational procedures for operating the system and any additional support that will be need.
For example, describe whether the equipment will be located in a TMC, at City Hall, at the corporation yard or signal shop and/or in the field. Will field equipment need to be field hardened or located within an air-conditioned environment? Will existing power supplies be adequate or will additional service, UPS and battery backups be required?
Will the operators be on duty or available 24/7 or during limited hours? Describe their required experience, skills and additional training needs.
Sample statements that may be used in this section are contained in the Concept of Operations samples table (Appendix B).
Chapter 7: Adaptive Support Environment
This chapter describes the current and planned physical support environment. Describe what support equipment, personnel, training and procedures currently existing, and explain those that need to be acquired or implemented.
Describe any additional test equipment and repair tools that will be needed to support the adaptive operation. Where will test equipment be located? Will system simulators be required (e.g., hardware and software to allow simulation of traffic and signal timing before new timing data goes live). Will a development server be required to set up controller firmware before deployment, and to test system upgrades and modifications before deployment?
Describe additional staff or contractors who will not be involved in the day to day operation of the system, but will be needed to support the operators and maintenance staff. This should include staff from the system vendor and/or consultants, who will provide additional on-going training, periodically audit the system setup and performance and support expansion of the system in the future.
Where multiple agencies are involved, describe the support that will be provided by or to other agencies. This should include any existing or proposed memoranda of understanding or operations and maintenance agreements that will affect the adaptive system, or will need to be modified to include reference to the adaptive system. This may include modifying the policies and procedures of those agencies in addition to developing new policies and procedures within your agency.
During the life of the adaptive system, will any equipment or supplies specifically related to the adaptive system need to be disposed of? Can this be done using existing procedures and protocols, or will additional arrangements be needed? At the end of useful life of the system, will any special disposal arrangements be required?
Sample statements that may be used in this chapter are contained in the Concept of Operations samples table.
Chapter 8: Proposed Operational Scenarios
Using an Adaptive System
The purpose of this chapter of the Concept of Operations document is to provide examples that illustrate how the system will be expected to operate and interface with the operators in typical circumstances. It is not intended to comprehensively describe the operation under all conditions. It is intended to illustrate to vendors, managers and decision-makers alike how you see your objectives being met by the system. This description is practically oriented and takes into account the practical limitations of available systems, which you expect to be live with. It should not be a description of how you would like some imagined system to operate with no regard for the practical limitation of candidate systems.
Each statement in a scenario should relate to a user need, although not all needs will be further described in a scenario. The statements in the description of each scenario do not directly generate requirements. Requirements are only generated by needs. The scenarios simply provide examples of how the system meets some of the needs.
Once you have written the scenarios, if you are not satisfied that they describe an operation that will be adequate, you should then review your needs statements. If you wish to describe elements of the proposed operation that are not described by needs, then additional needs should be enunciated.
8.1 How to Construct a Scenario
Each scenario should describe a unique set of circumstances, applying to one type of location, one set of traffic conditions with one set of appropriate activities by stakeholders, and one response by the ASCT system. A scenario should include statements about each of the following elements:
- Road network on which the scenario occurs
- Traffic conditions that must be accommodated
- The operational objectives that should be satisfied in this situation
- The coordination and signal timing strategies that should be applied by the ASCT system to satisfy those objectives.
If all the scenarios relate to the same section of road network, the description does not need to be repeated within each scenario. If the network is the same as described in the existing conditions, then the description may be deleted from the scenarios.
Example statements are included in the appendix. Each one is linked to an appropriate need. If you edit these statements or create your own, you must ensure that each new statement is matched to at least one of your statements of need, otherwise, you will not have a requirement that will ensure the system allows you to operate as you describe in your scenario.
Example scenarios used in a real project are also provided in Appendix C.