5.4 Actuated Coordinated Systems and Force-Off Modes

2.1 Actuated Coordinated Systems & Force-Off Modes

In modern traffic signal engineering, coordinating traffic signals along an arterial corridor or network is critical for reducing vehicle stops, minimizing delay, lowering fuel consumption, and decreasing vehicular emissions. While fixed-time signal coordination enforces strict, non-actuated green intervals regardless of real-time demand, actuated coordinated systems combine the bandwidth progression benefits of signal coordination with the local responsiveness of vehicle actuation. In an actuated coordinated system, specific phases are designated as coordinated (usually main-street through phases), while remaining phases operate on demand. Managing how and when non-coordinated phases are terminated requires precise control mechanisms known as force-offs, guided by permissive windows and a synchronized background cycle clock.

---

Principles of Actuated Signal Coordination

Actuated coordination relies on a common reference time—the background cycle length ($C$)—shared across all intersections along a corridor. Each local controller maintains an internal system clock synchronized to a master clock, GPS reference, or peer-to-peer network standard (such as NTP). The cycle is anchored by a designated Sync Reference Point (or local zero), which defines the starting position of the coordinated phase relative to adjacent intersections.

Coordinated Phases vs. Non-Coordinated Phases

Signal phases within an actuated coordinated controller are segregated into two distinct operational classes:

1. Coordinated Phases (Typically Phases 2 and 6): These phases serve the primary arterial through traffic. Coordinated phases are unique because they do not gap out or max out. Instead, they absorb all leftover or unallocated cycle time. They remain green until the designated Yield Point in the background cycle, guaranteeing minimum progression bandwidth for arterial platoons.
2. Non-Coordinated Phases (Phases 1, 3, 4, 5, 7, and 8): These phases serve side-street through movements and main-street or side-street left turns. Non-coordinated phases operate strictly on actuation. If no demand (vehicle or pedestrian call) exists when their permissive window opens, the controller skips them entirely, returning green time to the coordinated phases. If demand exists, non-coordinated phases terminate via gap-out (vehicle passage timer expires) or force-off (controller terminates green at a scheduled background time).

---

Coordinated Phase Yielding and Phase Splits

To maintain system synchronization, each phase is allocated a specific portion of the background cycle known as its Phase Split ($S_i$). A phase split is expressed either in seconds or as a percentage of the total cycle length ($C$), where:

$\sum S_i = C$

Each phase split includes the green interval, yellow change interval ($Y$), and red clearance interval ($R_c$). For non-coordinated phases, the maximum green time available is $Split - (Y + R_c)$.

The Yield Point represents the exact moment in the controller's background clock when the coordinated phase is permitted to yield the right-of-way to waiting non-coordinated calls. If no calls are registered on non-coordinated phases at the yield point, the controller remains in the coordinated green phase. If a call is placed after the yield point, servicing depends on whether the controller uses single or multiple permissive windows.

---

Force-Off Modes: Fixed vs. Floating Force-Offs

The Force-Off is a control signal generated by the coordination logic that terminates a non-coordinated phase regardless of vehicle presence or detector activity. Its primary purpose is to ensure that a non-coordinated phase does not exceed its allocated split time, thereby protecting downstream phases and preserving arterial progression bandwidth. Signal controllers implement two primary force-off operational modes: Fixed Force-Offs and Floating Force-Offs.

Fixed Force-Off Mode

In Fixed Force-Off mode, the force-off point for every non-coordinated phase is fixed to a specific, absolute point in the background cycle clock relative to the system sync point.

• Operation: If an upstream non-coordinated phase gaps out early (e.g., Phase 1 uses only 10 seconds of an allocated 20-second split), the unused 10 seconds become available to subsequent non-coordinated phases (e.g., Phase 2 or Phase 3) in the ring structure.
• Advantage: Subsequent non-coordinated phases can initiate green earlier than scheduled and run until their fixed force-off point, potentially receiving extra green time to clear unexpected queues.
• Impact on Coordination: Unused time cascades through the ring until absorbed by the coordinated phases. Fixed force-offs are highly effective in heavily congested corridors where side-street demand fluctuates significantly and extra green capacity is needed for non-coordinated movements.

Floating Force-Off Mode

In Floating Force-Off mode, the force-off point for a non-coordinated phase is dynamically calculated relative to when the phase actually begins its green interval.

• Operation: The force-off point "floats" with phase initiation. The maximum green duration for the phase is strictly capped at $Split - (Y + R_c)$, regardless of when the phase started green.
• Advantage: Unused green time from an early gap-out is not transferred to subsequent non-coordinated phases. Instead, any saved time is immediately returned to the coordinated phases.
• Impact on Coordination: Floating force-offs maximize green time dedicated to the arterial main street. They prevent side-street phases from expanding beyond their allocated split durations when preceding phases gap out early, maintaining tight control over arterial progression bands.

---

Permissive Windows and Yield Point Dynamics

A Permissive Window defines the time frame during the background cycle clock in which the local controller is authorized to yield right-of-way from the coordinated phase to service non-coordinated phase calls. Outside of this window, calls on non-coordinated phases are held over until the next cycle to prevent breaking coordination logic.

Single Permissive Window

In a Single Permissive Window architecture, only one opportunity exists per cycle for the controller to yield right-of-way. The permissive window begins at the Yield Point and remains open for a defined duration. If no call exists on a non-coordinated phase while the window is active, the window closes. Any vehicle arriving after the window closes must wait through the remainder of the current cycle and the coordinated phase before receiving a green indication in the subsequent cycle.

Yield Engine and Dynamic Permissives (NEMA TS 2 & 2070 Standards)

Modern NEMA TS 2 and Caltrans 2070 controllers utilize dynamic Yield Engines or multiple permissive windows. In these systems, each non-coordinated phase has its own individual permissive window calculated backward from its force-off point based on minimum green and clearance requirements:

$\text{PermissiveStart} = \text{ForceOffPoint} - G_{min} - Y - R_c$

This ensures that if a call is registered, the controller will only yield if sufficient time remains in the background cycle to service the minimum green and clearance intervals of the phase without violating downstream force-offs or the main-street sync point.

---

System Timing Calculations and Parameter Interdependencies

Designing actuated coordination plans requires calculating precise timing parameters to guarantee legal clearance intervals and prevent cycle failure.

1. Minimum Force-Off Calculation: The split allocated to any non-coordinated phase must be greater than or equal to the sum of its minimum green, yellow change, and red clearance: $S_i \ge G_{min,i} + Y_i + R_{c,i}$
2. Pedestrian Clearance Constraints: If pedestrians are permitted on a non-coordinated phase without resting in walk, the phase split must accommodate the pedestrian walk ($W$) and pedestrian clearance ($FDW$) intervals: $S_i \ge W_i + FDW_i + Y_i + R_{c,i}$ If the allocated split is less than the required pedestrian time, registering a pedestrian call will cause the controller to transition out of coordination (dwell/inhibit), forcing a temporary off-coordination cycle.