5.5 Time-of-Day (TOD) vs Traffic-Responsive Coordination

2.2 Time-of-Day (TOD) vs Traffic-Responsive Coordination

Signal coordination plans must adapt to shifting traffic patterns throughout the day and week. An arterial corridor that experiences heavy northbound commuter flow during the morning peak may see balanced off-peak traffic at mid-day and heavy southbound flow during the evening peak. To maintain optimal vehicular progression, traffic signal controllers utilize two primary system control strategies: Time-of-Day (TOD) pattern selection and Traffic-Responsive Program (TRP) control. While TOD relies on historical traffic data and fixed time schedules, TRP dynamically evaluates real-time detector data to select the most effective Cycle Length, Offset, and Split (COS) pattern from a pre-configured timing library.

---

Overview of Coordination Control Strategies

Signal timing patterns consist of three fundamental building blocks, collectively referred to as a COS Pattern:

1. Cycle Length ($C$): The total time required for one complete sequence of signal phases. Higher cycle lengths are selected during heavy traffic periods to increase intersection capacity by reducing the percentage of lost time dedicated to yellow and red clearance intervals.
2. Offset ($O$): The time difference (in seconds or percentage of cycle length) between the local intersection sync point and the system master reference zero. Proper offsets create a green wave that allows platoons of vehicles to travel through consecutive intersections without stopping.
3. Split ($S$): The percentage or duration of cycle time allocated to each individual phase at an intersection.

Selecting when to implement specific COS patterns determines corridor efficiency.

---

Time-of-Day (TOD) Scheduling and Transition Logic

Time-of-Day (TOD) control is a deterministic pattern selection method based on historical traffic counts. A TOD schedule consists of a clock-based calendar matrix that specifies pattern changes by hour, minute, day of the week, and holiday/season schedules.

Structure of TOD Tables

In a standard NEMA TS 2 or Type 2070 controller, TOD control operates through a multi-tiered database structure:

• Time Base Rules: Maps specific calendar dates and days of the week to Day Plans (e.g., Weekday, Saturday, Sunday, Holiday).
• Day Plans: Defines specific time-stamped trigger points throughout a 24-hour period (e.g., 06:30 AM -> Select Pattern 1; 09:00 AM -> Select Pattern 2).
• Pattern Tables: Maps pattern numbers (Pattern 1, Pattern 2, etc.) to specific Cycle, Offset, and Split (COS) configurations stored in memory.

Pattern Transition Smoothing Methods

When transitioning between two different coordination patterns (e.g., switching from a 90-second cycle to a 120-second cycle), the controller must adjust its internal background clock to align with the new offset and cycle length. Transitioning abruptly can disrupt traffic platoons and create safety hazards. Controllers execute transitions using three primary algorithm types:

---

Traffic-Responsive Plan (TRP) Selection and Algorithmic Logic

While TOD scheduling functions well under predictable, recurring traffic patterns, it cannot adapt to unscheduled events such as traffic incidents, inclement weather, construction detours, or special event surges. Traffic-Responsive Program (TRP) control bridges this gap by continuously monitoring real-time arterial traffic conditions via system detectors and automatically selecting the best pre-stored pattern from a library.

The System Metric: Volume plus Weighted Occupancy ($V + kO$)

TRP algorithms evaluate traffic demand across arterial corridors using a combined metric known as $V + kO$, which calculates weighted demand from key system detectors:

$\text{System Metric} = \text{Volume} + (k \times \text{Occupancy})$

Where:

• $\text{Volume}$ ($V$): Vehicle count per unit time (vehicles per lane per hour).
• $\text{Occupancy}$ ($O$): Percentage of time a detector zone is occupied by vehicles (0% to 100%).
• $k$ (Weighting Factor): A field-configurable weighting constant (typically ranging from 2 to 10) that scales occupancy relative to volume. Because occupancy surges dramatically during traffic congestion and queue spillback, $k$ ensures that density changes trigger appropriate pattern responses even when volume drops due to gridlock.

TRP Threshold Tables and Hysteresis

The system master evaluates smoothed $V + kO$ data against a threshold matrix. To select appropriate patterns, detectors are grouped into functional channels:

• Inbound Demand Channels: Monitor directional flow toward the central business district (CBD).
• Outbound Demand Channels: Monitor directional flow away from the CBD.
• Cross-Street / Circumferential Channels: Monitor side-street and network grid load.

To prevent rapid, erratic pattern toggling—a dangerous phenomenon known as pattern hunting—TRP algorithms enforce Hysteresis and Evaluation Filters:

1. Evaluation Period: Raw detector data is smoothed using exponential rolling averages (typically over a 5 to 15-minute sampling window) to filter out momentary spikes caused by single platoons.
2. Hysteresis Threshold Offsets: The threshold required to transition up to a higher-capacity pattern (e.g., Pattern 1 to Pattern 2 at $V+kO = 75%$) is set higher than the threshold required to drop back down (e.g., Pattern 2 to Pattern 1 at $V+kO = 65%$). This 10% hysteresis buffer prevents continuous pattern switching when demand hovers near a boundary.

---

System Detector Placement, Filtering, and Maintenance

The accuracy of TRP control depends directly on correct System Detector placement and configuration. System detectors differ functionally from local intersection stop-bar detectors.

Placement Guidelines

• Location: System detectors are placed mid-block, typically 1/3 to 1/2 mile downstream from the nearest upstream signalized intersection (or 200–500 feet past major commercial driveways and bottlenecks). This isolates detectors from stop-bar queue spillback and ensures they measure free-flowing platoon volume and true corridor density.
• Lane Coverage: System detectors should cover every lane of key approach directions to capture total multi-lane volume and average occupancy.

Fail-Safe Operational Strategies

System detectors operate in harsh roadway environments and are subject to hardware failures (e.g., severed inductive loops, misaligned radar sensors). TRP controllers implement robust diagnostics:

• Error Detection: If a system detector reports constant call (stuck on), zero counts over extended peak periods (stuck off), or erratic chatter, the controller flags the detector as invalid.
• Detector Re-weighting & TOD Failover: If a primary system detector fails, the TRP engine re-weights remaining active detectors. If critical detector channels drop below minimum operational thresholds, the system automatically falls back to the background Time-of-Day (TOD) schedule, generating an administrative alert to field technicians.