Queue Warning

CTCAV-046
Vehicle SafetyMobilitySafety
TRL 7
Overview

Queue Warning uses existing road sensing systems, connected vehicle data, and Vehicle-to-Everything (V2X) communications to provide real-time queue detection and share queue information and warnings with nearby upstream vehicles and centers, such as the Traffic Management Center (TMC). This application does not function as a crash avoidance system. Instead, it aims to provide timely alerts about potential crash situations to the driver, reducing the need for crash avoidance or mitigation actions later.

Involved Centers

Caltrans Centers:

TMC

Other Centers:

TIC
Devices

RSU

OBU

V2X Communications
Communication methods used
V2IV2N
Deployment Maturity
Technology Readiness Level
7

Prototype demonstrated in operational environment

Development
TRL 1TRL 9
ARC-IT Reference
Architecture Reference for Cooperative and Intelligent Transportation

Deployment Specification

v1.0 Draft
Objectives

Queue Warning uses existing road sensing systems, connected vehicle data, and Vehicle-to-Everything (V2X) communications to provide real-time queue detection and share queue information and warnings with nearby upstream vehicles and centers, such as the Traffic Management Center (TMC). This application does not function as a crash avoidance system. Instead, it aims to provide timely alerts about potential crash situations to the driver, reducing the need for crash avoidance or mitigation actions later. This application involves two main tasks: queue determination (detection and/or prediction) and sharing queue information through vehicle-based, infrastructure-based, or hybrid methods.

Expected Benefits

This application aims to improve safety for both motorized and non-motorized road users. It also aims to enhance transportation system efficiency by reducing delays, fuel consumption, and emissions. Additionally, it supports effective system management and operations, boosting the resilience and reliability of the transportation network.

Performance Measures

Safety

Number of crashes, injuries, and fatalities for a selected time period, such as a month or a year.

Mobility

  • Point-based or section-based flow rates (or traffic counts) and speeds (e.g., average, 90th/95th percentile, and variance)
  • Section-level or network-level delays, energy/fuel consumptions, travel times (e.g., average, 90th/95th percentile, and variance)

Caltrans Implementation Options

There are two main tasks in Queue Warning: queue detection and queue warning dissemination. Global queue detection at the TMC is recommended to better evaluate network traffic conditions. For dissemination, both direct communications (via RSUs) and network communications (via cellular providers) can be used.

OptionQueue Detection MethodDissemination Method
Option AAt TMCNetwork Communications
Option BAt TMCDirect Communications
Option CAt TMCDirect & Network Communications
Global Queue Detection and Network Communications
This option uses global queue detection at the TMC and disseminates queue warnings via network communications (e.g., through cellular service providers like AT&T). Since Queue Warning is not a latency-critical application, network communications are more suitable because they don't require installing RSUs at road intersections and can cover a larger area for message dissemination.

Physical architecture for the deployment option of global queue detection and network communications. Click the diagram to enlarge.

System Components
Key players, devices, and required functions

Data Flows & Standards
Data FlowContentStandard
(1) Detector DataTraffic flow measures like vehicle counts, occupancy, and speed.TMDD (optional)
(2) Queue DataDetected vehicle queues on roadways with information like time, location, severity, etc.Vendor Specific
(3A) BSM MessagesInformation that includes vehicle location, heading, speed, acceleration, brake status, etc.SAE J2735
(3B) Aggregated CV DataAggregated BSM messages in CSV files.SAE J2735
(4) Third-Party DataReal-time or near-real-time traffic incident alerts, traffic flow data, and road closures.Various Formats
(5) Collected Traffic DataData collected from various sources, including loop detector data, queue data from advanced sensors, CV data from CV application providers, and incident and traffic data from third-party applications.Various Formats
(6A & 6B) Queue Warning MessagesTraveler Information Messages (TIMs) with information regarding formed or impending queues (location, estimated duration, and other descriptions) and recommendations, including speed reduction, lane change, or reroute.SAE J2735

Implementation Considerations

Data Sources for Queue Detection

Loop Detector Data
Coverage
High

Inductive loop detectors are widely used in California, especially on major freeways. However, the statewide detector health rate is low, falling below 50%.

Granularity
Low

Data from individual detectors is point-based and usually aggregated over five-minute periods.

Accuracy
Moderate

With average spacing of 1 mile between stations and shockwave speed of 15 mph, detection delay stays within 5 minutes and end-of-queue error within 1 mile. Reliability is high as detectors operate 24/7 regardless of conditions.

Cost
Low Deploy / High Maintain

Deployment cost is minimal due to extensive existing coverage. Maintenance costs are high because statewide detector health is below 50%.

Advanced Traffic Detection Sensors
Coverage
Low

Currently limited, but expected to expand rapidly as more infrastructure upgrades incorporate these advanced sensors.

Granularity
High

Individual object movements can be tracked within the detection zone in real time or near real time.

Accuracy
High

High accuracy under normal conditions. Performance may decline in adverse weather, but queue detection is expected to stay accurate since only high-level queue information is required.

Cost
High

Each sensor typically costs a few thousand dollars for procurement and a few hundred to a thousand dollars annually for data management.

Connected Vehicle Data
Coverage
Network-level

Unlike point-based loop detectors and section-based sensors, CV data coverage occurs at the network level.

Granularity
High

Derived from BSM data including vehicle location, heading, speed, acceleration, brake status, and more. Quality depends heavily on CV penetration rate.

Accuracy
Low (currently)

Challenging due to low penetration rates. Useful as additional data source combined with loop detector data. Once penetration reaches ~10%, may offer accurate and reliable detection alone.

Cost
Varies

High when collected from RSUs, more affordable when from V2N providers like AT&T and Verizon.

Third-Party Traffic Data
Coverage
High

Extensive coverage from multiple sources such as application users and vehicle fleets.

Granularity
Moderate

Often aggregated due to low penetration rates of consumers and vehicle fleets.

Accuracy
Variable

Depends on quality of third-party data, linked to consumer/fleet penetration rates and fusion algorithms. Currently a black box to users.

Cost
Moderate

Varies depending on project area and subscription fees. Doesn't require major infrastructure upgrades.

Queue Warning Settings

Queue Warning Zone

When the end of a queue is detected, Traveler Information Messages (TIMs) will be generated in the Queue Warning System within the TMC and sent to the RSUs or V2N Connected Vehicle Application Providers to alert connected vehicles upstream of the queue. An essential step is to define a local queue warning zone where connected vehicles receive the TIMs and take appropriate actions.

A longer warning zone is necessary when:

  • Severe downstream traffic congestion due to work zones, traffic accidents, special events, and other factors
  • Heavy upstream traffic demand
  • Detour options are included in the TIMs
Queue Warning Suggestions
Recommendations included in TIMs to help connected vehicles take appropriate actions

Speed Limit

Setting a lower speed limit can slow down upstream vehicles, potentially creating more space for queued vehicles to move faster. This may improve safety for vehicles approaching the end of the queue.

Consideration

Speed limits should be chosen carefully. Choosing a lower speed limit than necessary can cause significant delays upstream, offsetting the time savings downstream.

Lane Change

Advising connected vehicles to change lanes earlier before reaching the end of the queue.

Consideration

If the lane change advisory is issued too close to the end of the queue, connected vehicles may not have enough time to make timely lane changes and might overreact, disrupting the entire traffic flow.

Detour Route

For severe downstream congestion with long queues, detour routes can be provided to connected vehicles further upstream, redirecting traffic to local arterials.

Consideration

Closely related to Integrated Corridor Management (ICM). Detour routes should be selected carefully, and timing plans need to be updated. The use of detour routes should be monitored in real time to ensure they do not become overwhelmed.

Testing & Performance Evaluation

Testing & Verification

Device / Component Level

For each device or component, test and verify: (i) whether the required data standards are correctly applied; (ii) whether data inputs can be properly read; (iii) whether it can perform the necessary functions; and (iv) whether it can produce outputs in the correct format with accurate information.

System Level

End-to-end testing to verify that: (i) messages or data files are transmitted accurately between components or devices; (ii) all components and devices support the application, from detecting queues from various sources, to generating queue warning messages, and to sending those messages to upstream connected vehicles.

Performance Evaluation

Microsimulation (Pre-Deployment)

PATH has developed a V2X microsimulation platform capable of thoroughly evaluating the queue warning application across data availability, CV penetration rates, communications range/latency/packet loss, congestion severity, traffic demands, recommendations, zone settings, and driver reactions.

Before-and-After Analysis (Post-Deployment)

Once deployed, data from detectors, advanced sensors, connected vehicles, and third-party applications can support post-analysis. Reductions in crashes, injuries, and fatalities serve as safety improvement measures. Flow, speed, delays, energy consumption, and travel times serve as mobility enhancement indicators.