Office of Operations
21st Century Operations Using 21st Century Technologies

Traffic Control Systems Handbook: Chapter 8. System Control

Photograph of two employees monitoring traffic on a number of CCTV displays and computer terminals at a TMC.
Figure 8-1. Traffic Management Center (TMC) (1).

8.1 Introduction

Section 3.8 describes functional categories for system control of traffic signals. This chapter concentrates on the architectures and the physical implementation of these categories. Table 8-1 briefly identifies these categories and their major architecture characteristics.

Table 8-1. Performance Levels for Traffic Signal Systems
Category Of System Operation Architecture Characteristics
Traffic Adaptive Control
  • Cycle free, rapid reaction to sensed traffic conditions
  • Distributed Control
  • One or two detectors per signalized approach
Traffic Responsive Control
  • Rapid reaction to sensed traffic conditions
  • Central control (SCOOT) or distributed control (SCATS)
  • One or two detectors per signalized approach
Traffic Adjusted Control
  • Area traffic adjusted control.
  • Critical intersection control (centralized architecture only).
  • Local intersection strategies.
  • Interconnection required
  • Moderate number of system detectors required for traffic responsive timing plan selection
  • Three distributed computation levels (closed loop) or
  • Two distributed computation levels or
  • Central control
Interconnected Control
  • Time of day or operator selected timing plans.
  • Local intersection strategies.
  • Interconnection required
  • No system detectors required for timing plan selection (time of day or operator selection only)
  • Three distributed computation levels (closed loop) or
  • Two distributed computation levels or
  • Central control
Time Base Coordination
  • Time of day plans.
  • Local intersection strategies.
  • Provides basic coordination.
  • No interconnection required
Uncoordinated Signals
  • No coordination among traffic signals

8.2 Architectures for Conventionally Coordinated Traffic Control Systems

Signal timing plans for the lower four levels of Table 8-1 are stored in the controller's database. Interconnected Control and Traffic Adjusted Control require interconnection by means of wireline or wireless techniques and provide the capability for time of day operation and operator selection of signal timing plans. The following architecture variations may be used to implement these two levels.

  • Three Distributed Computation Level Architecture (also known as "closed loop" systems).
  • Two Distributed Computation Level Architecture.
  • Central Control Architecture.

Three Distributed Computation Level Architecture

Figure 8-2 shows how the computation, control and equipment functions are distributed among the three levels. The TMC equipment complement consists of servers, communication equipment, local area network, and peripheral equipment such as printers. The database for the system is stored in the servers and downloaded to the field master controller and then to the local intersection controllers.

Flow diagram describing architecture for a typical three computation level distributed system.
Figure 8-2. Typical Architecture for Three Computation Level Distributed (Closed Loop) System

The architecture contains one or more field master controllers. These units may be located in controller cabinets in the field or may be located at the TMC. Their function is to break down the information from the TMC into the communication channels appropriate to each local intersection controller, and to assemble data from these controllers into an information stream for transmission to the TMC. In the case where the system is traffic adjusted, system detector data from the local field controllers is processed in the field master controller and is used to select the appropriate timing plan. The traffic responsive control algorithm for closed loop systems is described in Section 3.8.

The local intersection controller controls the traffic signal displays. It stores the library of timing plans available for control and implements the selected timing plan (the information for the selection of traffic responsive control timing plans originates at the field master controller and the information for operator selection originates at the TMC). Time of day control uses schedules downloaded from the TMC and stored in the local intersection controller.

The local intersection controller preprocesses the system detector data into volume and occupancy parameters that are uploaded to the higher levels at intervals that might range from several seconds to a minute. Based on data received from the local detectors, the local controller terminates actuated phases at the appropriate time.

Three level systems are often provided by traffic equipment suppliers.

Two Distributed Computation Level Architecture

The functions for this architecture are identical to those identified for the three distributed computation level architecture discussed above; however, all of the functions performed by the field master controller are performed by the TMC equipment complement. The functions of the local field controller remain unchanged. Two level systems are often provided by traffic system software suppliers.

Central Control Architecture

Figure 8-3 shows the signal flow for this architecture, which is currently used less frequently than in the past. While the TMC performs all of the computation functions as for the previously described architectures, it also performs the following major functions:

  • Processes and smoothes system detector data.
  • Uses system detector data to select timing plans based on a traffic responsive control algorithm. Section 3.8 describes the First Generation UTCS Control Algorithm used by many systems.
  • Converts the timing plan to controller coordination commands and provides these commands to the field controller at precisely the proper time in the signal control cycle. Table 8-2 describes the functions of the coordination commands.

Flow diagram showing typical central control architecture.
Figure 8-3. Typical Arrangement for Central Control Architecture

Table 8-2. Coordination Commands
Name Function
Force-off Issued to begin termination of an actuated phase. Each assigned force-off point is located precisely in the background cycle. It is sent to the controller only if:
  • Its respective phase is called and serviced,
  • Calls on the phase continue to (and beyond) the force-off point, or
  • Calls are present on any of the subsequent scheduled phases.
The last (actuated) phase force-off occurs only if the above listed conditions are met. If calls on any of the actuated phases terminate before the scheduled occurrence of the force-off and demand is present on the next phase, the demanding phase will be served earlier than scheduled in the background cycle time frame.
Hold Issued to guarantee a certain minimum of green time to a phase. Normally used in the coordinated phase to guarantee a window of time in the background cycle for traffic progression on the major street. Thus, the coordination phase is considered non-actuated because its duration is not reliant upon vehicle calls. A hold command can be applied to a phase other than the coordinated phase.
Yield By de-energizing the Hold command, the resulting yield condition begins the termination of the active phase. The yield condition, no longer recognized by NEMA as a command, is usually associated with the coordinated phase, and in the absence of demand on other (actuated) phases, the controller remains in coordinated phase green. The yield (or permissive) condition may be programmed for a certain duration. If a call subsequently appears on any of the other (actuated) phases during the yield condition, the calling phase would, conditionally, be serviced. The condition for service is that sufficient time remains before the calling phase's scheduled force-off command occurs.

A remote communication unit (RCU) or, in some cases, an intelligent remote communication unit (IRCU) may be used to establish communication between the TMC and the intersection controller. Its information processing functions are generally minimal, and it primarily serves to match the interfaces and protocols required by the field controller to those used for communication with the TMC.

Under normal conditions, the computation functions in the local intersection controller are minimal. Commands from the TMC at precisely the proper time in the traffic signal cycle terminate the green interval for each phase. Indication of the state of the green interval is transmitted to the TMC for monitoring purposes. Backup timing plans are also stored in the local intersection controller to be used in case the TMC fails or communication is lost.

Since this architecture directly controls the phases (or in some cases the intervals) of the traffic cycle from the TMC, it can, in principle, facilitate the use of control strategies that may not be based on the use of stored timing plans such as certain types of transit priority strategies.

8.3 Traffic Responsive and Traffic Adaptive Systems

The conceptual aspects of these signal systems are discussed in Section 3.8. This section discusses the physical aspects of representative adaptive control systems as well as some of the performance tests.

Conventional traffic signal systems were discussed in Sections 8.2 and 8.3. These systems represent the largest number of systems currently in use in the United States. Timing plans for these systems are stored in the local controller, and may be selected from the traffic management center or by a field master controller in the following ways by:

  • Time of day schedule.
  • The operator.
  • A traffic responsive algorithm. The traffic responsive algorithms, as discussed in Chapter 3, select timing plans for a section based on filtered real-time traffic data. Response times are usually in the order of a few minutes, and these timing plans are typically employed for a significant period of time (commonly exceeding one half hour). Since systems of this type implement timing plans that are prestored in the controller, the transportation management center or field master controller cannot alter the signal timing on an intracycle (phase or interval) basis.

Since the intent of traffic responsive and traffic adaptive systems is to rapidly respond to changes in traffic flow by analyzing the measured flow of upstream vehicles, the required intracycle information (from the traffic management center or in some cases from adjacent intersections) must be appropriately processed by the controller software. Thus, standard NEMA TS2 controllers generally require software modification to operate with adaptive systems. Type 2070 and ATC controllers with appropriate software are often used for adaptive systems. More intensive deployment of traffic detectors is generally required for adaptive systems as compared with conventional traffic responsive systems.

A wide range of performance improvements have been reported in the literature for these systems. While test techniques vary, the baseline for these tests has generally been pretimed systems, with considerable variation in the quality of the time of day plans. Table 8-3 provides a representative sample of results obtained in this way. The greatest benefits occur for those situations where traffic experiences significant variation from a pretimed plan. Examples include special events and arterials that support diversion from freeways during incidents.

Table 8-3. Representative Results of Performance Testing
Adaptive Control Algorithm Test Location Results
OPAC Northern Virginia 5-6% improvement in stops and delays1
SCATS Oakland County, Michigan Average of 7.8 % reduction in delay2
Newark, Delaware Travel time reduction up to 25%2
SCOOT Minneapolis, Minnesota Delay reduction of up to 19% during special events2
Toronto, Ontario 8% decrease in travel time, 17% decrease in delay3
Durban, South Africa 7% travel time reduction4
1 Gartner, N.H., F.J. Poorhan, and C.M. Andrews. “Implementations and Field Testing of the OPAC Adaptive Control Strategy in RT-TRACS.” Presented at the 81st Annual Meeting of the Transportation Research Board, Washington, DC, 2002.
2 Sussman, J. “What Have We learned About Intelligent Transportation Systems.” Chapter 3, Federal Highway Administration Report No. FHWA-OP-01-006, December 2000.
3 “SCOOT in Toronto.” Traffic Technology, Spring 1995.
4 “Durban SCOOT System.” The Durban Metro Council, Durban, South Africa, February 27, 1996.

Table 8-4 identifies some advantages and disadvantages of traffic responsive and adaptive systems.

Table 8-4. Advantages and Disadvantages of Traffic Responsive and Traffic Adaptive Control Systems
  • May be significant advantages in stops, delays and emissions compared to pretimed systems (Level 2). Probably somewhat less improvement compared to conventional traffic responsive systems (Level 3).
  • No need to periodically update timing plans.
  • Higher initial cost for both field equipment and traffic management center software. Higher maintenance cost for field components.
  • More difficult initial system setup and tuning process.

ACS-Lite, a FHWA research project, adapts certain principles developed during adaptive control system research and development to use by closed loop systems. This discussion is adapted from Reference 2.

ACS-Lite employs the concept that a TOD schedule is an appropriate way to manage traffic demand over the day and by days of the week. Within the context of a TOD schedule, ACS-Lite will adapt the particular plans that are implemented at each time of day based on the overall performance of that plan for the similar previous day. This approach to adaptive behavior uses the traditional traffic engineering assumption that average behavior of traffic on, for example, Tuesday at 3 P.M., is roughly the same on every Tuesday at 3 P.M., but drifts slowly with long-term changes in population, construction, new routes, etc.

If the performance of the baseline plan is determined to be improvable by changing cycle, splits, or offsets, then those changes will be made to the "optimized" plan stored in ACS-Lite and downloaded to the local controllers for use on the next day. The goal of being appropriately adaptive at this level is the maintenance of the timing plan over long periods of time to address the typical degradation of plan effectiveness (e.g., 4% worse per year) and replace the very expensive task of re-timing signals on a periodic basis.

The next level of adaptivity used by ACS-Lite is on-line modification of the TOD plan parameters as the plan is running. With the assumption that the baseline optimized TOD plan is a good starting point, ACS-Lite will adapt the cycle, split, and offsets of the plan within some neighborhood of the baseline settings over the plan's intended implementation duration. ACS-Lite may also adapt the start and end time of the plan from the baseline TOD schedule according to the current conditions, considering the effectiveness of the new plan versus the one that is currently running. ACS-Lite also identifies and selects the best strategy to transition between timing plans.

8.4 Time Base Coordination

Signal coordination requires a common time reference shared by all controllers in the coordination group. They must each reference their offset to the same background cycle - a background cycle that is of the same duration and starts at the same time at all controllers. This can be achieved by a master controller transmitting a synchronization pulse or message to all the controllers at the start of the background cycle. However, such schemes fail if the communications link breaks down.

Modern controllers most commonly use time base coordination as the means of synchronizing the start of the background cycle in all controllers. This scheme uses a time-of-day clock in each controller. The clock enables each controller to know the current time to at least the nearest second. The controller considers the background cycle to have started at a particular time of the day, such as midnight, called the offset reference time. At any time during the day, the controller can determine where it is now in the background cycle by calculating the number of seconds since the offset reference time (say midnight) and dividing by the cycle length. The remainder is used to calculate when the current background cycle started. The local cycle zero point is then calculated by adding the offset time to the background cycle zero point.

Time-base coordination works regardless of the mix of controller types and software in the system. However, it only works if the clocks in all controllers are well synchronized. Each controller counts the passage of time and automatically adjusts for daylight saving time. Clocks in controllers tend to drift over time and need to be reset periodically. If the controller is connected to a master or central computer, the clock can be reset automatically.

Standard time can be obtained by computers, masters and laptops computers used by technicians using transmissions from various sources such as cellular phone systems, radio broadcast from the National Institute of Standards and Technology (radio stations WWV and WWVH), time servers on the Internet, and the Global Positioning System. Some agencies have installed wireless time-source receivers in each signal cabinet. Section 7.5 provides additional information on time base references.

8.5 Traffic Management Centers

A traffic management center (TMC) is the hub of a traffic control system. The TMC brings together human and technological components from various agencies to perform a variety of functions. TMCs may deal with freeway traffic management, surface street traffic management, transit management or some combination of these functions. Chapter 14 of the Freeway Management and Operations Handbook (3) gives detailed information on freeway TMCs. This section is concerned primarily with traffic signal system TMCs.

Functions of Surface Street TMCs

Typically, surface street TMCs perform the following functions (3, 4):

  • Implement dynamic selection of traffic signal timings
  • Implement transit signal priority
  • Provide coordination among various agencies
  • Monitor traffic signal equipment, and dispatch resources to fix malfunctioning equipment
  • Provide traffic detection and surveillance
  • Modify arterial traffic signal timing when an incident occurs on a freeway
  • Manage incidents and special events or emergency evacuations
  • Store data for long term archives and offline operation

Traffic Signal Timing

One of the primary purposes of a traffic signal system TMC is to manage the timing of traffic signals in urban networks and on arterial streets. Special software allows an operator at a workstation in the TMC to communicate directly with field equipment and modify traffic signal programs in real time.

One of the largest and most advanced traffic signal system TMCs is the Automated Traffic Surveillance and Control System (ATSAC) in Los Angeles, California. ATSAC has been upgraded to use the Adaptive Traffic Control System (ATCS). ATCS can operate in the following modes (5):

  • Adaptive: each intersection within a group operates on a common cycle length, determined by traffic conditions within the group. Local intersection splits are based on traffic demand at the intersection, and offset is selected to minimize number of stops on the approach link with highest flow and based on intersection separation distances.
  • Time-of-day: each intersection operates on a predefined fixed program, as selected by the traffic engineer, based on the time of day.
  • Operator Control: operator can override the normal program by selecting a special timing plan from the TMC. This is useful during special events and incidents.

A traffic signal system TMC is useful when helping to manage incidents. CCTV cameras, as well as other forms of surveillance, can be used to detect incidents. When an incident occurs, either an operator can manually select an incident traffic signal timing plan, or by the use of a traffic responsive expert system which automatically selects an incident timing plan. Changeable message signs, web sites and the media may be used to alert motorists.

Size and Staffing Requirements of TMCs

The size and staffing requirement of a TMC can vary greatly, depending on the size of the urban area, and the functions that the TMC provides. Table 8-5, below, shows a summary of TMC size and staffing requirements for various TMC's in the US (6, 7, 8). Since TMC functions differ among agencies, the site and staff size may not be comparable among these agencies.

Table 8-5. Comparison of Various Traffic Signal System TMCs in the United States
Location City / Area Approx. Population TMC Size Number of Traffic Signals on System Staff
Los Angeles, CA (ATSAC) 3,700,000 5500 sq ft 2912 7 transportation engineers, including 1 supervisor. 2 systems analysts, 1 graphics designer, 1 traffic signal electrician, 1 secretary
Miami-Dade County, FL 2,200,000 5000 sq ft 2020 13 employees
San Antonio, TX 1,100,000 6000 sq ft 765 1 engineer, 3 technicians
Las Vegas, NV: Las Vegas Area Computer Traffic System (LVACTS) 1,500,000 (Covers Clark County) 2500 sq ft 700 4 administrative positions. 4 traffic operations positions. 4 maintenance positions
Atlanta, GA 416,000 2300 sq ft 650 Traffic signal operations: 1 engineer, 1 senior operator, 2 operators. CCTV: 1 engineer, 1 technician
Albuquerque, NM 449,000 800 sq ft 450 4 employees (2 engineers)
Denver, CO 555,000 2800 sq ft 450 No dedicated staff for TMC. Approx 1.5 FTE, more during special events
Seattle, WA 600,000 1420 sq ft 432 One supervisor, two operators
Phoenix, AZ 1,300,000 1500 sq ft 400 1 supervisor, 4 technicians
Boston, MA 590,000 2500 sq ft 320 7-8 employees
Renton, WA 53,000 700 sq ft 96 Initially, one part-time staff member. Can accommodate up to two full-time staff members
Redmond, WA 48,000 800-1400 sq ft (currently under construction) for traffic management area. 1200-1700 sq ft for signal shop area 25 (under construction)

Control room: one supervisor, one operator

Signal shop: up to five maintenance staff

Sources: • Various TMCs • Freund, K. “A Comparison of Traffic Management Centers in the Puget Sound Region.” ITE Journal. Institute of Transportation Engineers, Washington, DC, July 2003, pp. 46-50. • United States Census Bureau. United State Department of Commerce. 2000. • Federal Highway Administration, USDOT, Oak Ridge National Laboratory, “Metropolitan Intelligent Transportation Systems Infrastructure Deployment Tracking Database — FY96.” January 1996.

Coordination between Traffic Systems and Traffic Agencies

Because a signalized arterial may cross the boundaries of traffic systems or jurisdictions, it may be necessary to coordinate the operation of the systems so that traffic progression is maintained across a boundary. This requires the following:

  • Identical cycle length on opposite sides of the boundary.
  • Establishment of appropriate offset between signals on opposite sides of the boundary.
  • Use of a common time reference by the two systems.
  • Coordination with maintenance in case of failure.

This type of coordination is facilitated by the use of common time of day schedules for the traffic systems on opposite sides of the boundary.

Where the systems lie in different jurisdictions, a memorandum of understanding between the agencies may be appropriate to establish the basis for a joint planning effort. Inter-agency coordination is discussed in Reference 9.

1. "Goals for a Livable City." City and County of Honolulu (

2. "ACS-Lite Algorithms Detail." Gardner Transportation Systems, November 19, 2002.

3. Neudorff, L.G., J.E. Randall, R. Reiss, and R. Gordon. "Freeway Management and Operations Handbook." Federal Highway Administration Report No. FHWA-OP-04-003, Washington, DC, September 2003.

4. "Transportation Management Center Concepts of Operations: Implementation Guide." Federal Highway Administration Report No. FHWA-OP-99-028 / FTA-TRI-11-99-23, Washington, DC, December 1999.

5. "ATCS: Adaptive Traffic Control System." City of Los Angeles Department of Transportation, April, 1998.

6. Freund, K. "A Comparison of Traffic Management Centers in the Puget Sound Region." ITE Journal. Institute of Transportation Engineers, Washington, DC, July 2003, pp. 46-50.

7. United States Census Bureau. United State Department of Commerce. 2000.

8. Federal Highway Administration, USDOT, Oak Ridge National Laboratory, U.S. Department of Transportation. "Metropolitan Intelligent Transportation Systems Infrastructure Deployment Tracking Database - FY96." January 1996.

9. "Cross-Jurisdictional System Coordination Case Studies." Science Applications International Corporation, Federal Highway Administration Research Report No. FHWA OP-02-034, February 2002.

Next | Previous

Office of Operations