NCDOT, Vizzion, and Iteris Join Forces to Improve Driver Safety and Traffic Flow with On-Vehicle Camera Data

In 2023 a collaborative effort between the North Carolina Department of Transportation (NCDOT), Vizzion, and Iteris leveraged the power of on-vehicle camera data for enhanced traffic management.

Traffic Management in the Age of Data

As traffic on our roads continues to grow, data-driven solutions are becoming increasingly important for effective management. This research delves into the potential of on-vehicle camera images to provide real-time insights into traffic incidents, which can improve response times and situational awareness for transportation agencies.

Validating Existing Data and Filling the Gaps

The study investigates how on-vehicle camera images can be integrated with existing traffic management frameworks. By comparing these images with user-submitted fog data via Waze and logged crash and roadwork data, the research assesses the accuracy of current information and identifies areas where on-vehicle cameras can provide valuable additional perspectives. This can help to improve the overall quality of traffic data and fill in gaps in areas not covered by traditional stationary cameras and sensors.

Vizzion's Contribution: Real-Time Views from the Road

Vizzion's Drives service plays a key role in this project. It offers live views and historical footage (up to 10 days back) captured by cameras installed on commercial vehicles throughout Vizzion's network. With these cameras capturing images every second while the vehicle is moving, the data provides a highly granular view of the road, with images roughly every 30 meters or less. The network currently has around 60,000 active vehicles, collectively traveling over 5 million miles each day.

"On-vehicle camera data like Vizzion's, and its potential use within a system like the ClearGuide platform, promises a revolutionary way of seeing things on the roadway," says Kelly Wells, PE, NCDOT Traveler Information Engineer. "I think this has the potential to be the best thing since probe speed data to improve situational awareness for traffic incident management and traveler information."

Benefits of On-Vehicle Camera Technology

The use of on-vehicle camera data offers several advantages for traffic management:

• Enhanced Situational Awareness: Real-time views from on-vehicle cameras can provide a more complete picture of what is happening on the road, especially in remote locations or areas not covered by CCTV cameras.
• Improved Incident Response: Faster detection and verification of incidents can lead to quicker response times, reducing congestion and improving safety.
• Data Verification: On-vehicle camera images can be used to validate existing data sources, ensuring the accuracy of traffic and weather information.
• Cost Savings: By providing real-time views, on-vehicle camera data can potentially reduce the need for agencies to dispatch personnel to investigate incidents or install costly additional CCTV cameras and weather sensors.

The TIMS Pilot Study: NCDOT, Vizzion, and Iteris Working Together

This pilot study specifically explored the use of Vizzion's on-vehicle camera data to verify the accuracy of crash and roadwork incidents logged in NCDOT's Traveler Information Management System (TIMS). The study focused on four freeway segments in North Carolina with known active work zones.

The researchers compared logged incidents with available on-vehicle camera images and identified additional incidents not previously recorded in TIMS. This demonstrates the potential of on-vehicle camera data to enhance NCDOT's awareness of traffic events and improve the quality of its data.

In this study, Vizzion's on-vehicle camera images were investigated to check the accuracy of the logged TIMS crash and road work incidents.

Freeway locations with active work zones were considered. Based on a review of October 2023 monthly congestion contour plots from ClearGuide for frequent traffic, four bi-directional segments were selected: one on I-40 and three on I-95, as listed and shown in the map below. All segments were in North Carolina.

1. I-40 Mile Marker (MM) 283-289 in Wake County
2. I-95 MM 13-22 in Robeson County
3. I-95 MM 56-71 in Cumberland and Harnett Counties
4. I-95 MM 71-81 in Harnett and Johnston Counties

Combined, the segments have a total length of 40 centerline miles.

Two steps were followed for the review. First, for the known TIMS incidents (road work or crashes), on-vehicle camera images were queried, if available. Second, all images were viewed and additional incidents not logged in TIMS were noted.

The figures below show examples of on-vehicle camera images of crash and roadwork incidents, respectively.

On-vehicle camera image of a crash. 11/17/2023 on I-95 Southbound in Harnett County.

On-vehicle camera image of roadwork. 11/18/2023 on I-95 Northbound in Harnett County.

Nearly half of the logged crash and road work incidents were confirmed through inspection of on-vehicle camera images. Additionally, six previously undocumented work zones were found through Vizzion's imagery in the one weeklong study. Discussion of the availability of images (market penetration ) and its effect on incident coverage was provided. Overall, the work in this paper has shown that the use of near real-time on-vehicle camera images can enable a transportation agency to have enhanced awareness of incidents, as well as improve the quality of logged incident data.

Additional work on incident verification could include detailed checks of the locations and start/end times of logged road work incidents, as well as subsequent corrections to the TIMS incident data. The development of automated tools and UI for searching and combining traffic incidents and on-vehicle camera image data could be used to support traffic management and analysis. A mockup of a tool for TMC operators to view real-time on-vehicle camera images is shown below.

Mockup of real-time on-vehicle camera images integrated into ClearGuide.

The Visibility Pilot Study : Confirming Fog and Unique Events

In an extension of the pilot study, over a one month period, NCDOT requested on-vehicle imagery near each fog alert received through Waze around Boone, NC. Of the 28 fog alerts meeting the criteria of the pilot, Vizzion's on-vehicle imagery was able to confirm and provide additional confidence and context for eight of the events. The imagery provides a sense of severity of the incident; NCDOT was able to classify the fog events from the imagery as low, medium, and high fog events, providing additional information that can impact the DOT's response.

In one out of the eight events with imagery, no evidence of fog was found. This can enable the agency to disregard the event as false. If NCDOT decides to implement fog alerts for travelers, this application would save unnecessary response efforts and avoiding false information being provided to the driving public.

This additional confidence and context allow DOTs to take actions such as updating variable speed limits and alerting drivers through Variable Message Signs (VMS), 511 sites, and familiar consumer channels and apps such as in-cab alerts. The ability to use crowdsourced data that is confirmed by on-vehicle imagery provides a lower cost solution than an RWIS Fog Detector and is more likely to be selected for implementation due to a better return on investment.

During the pilot, there was an incident on I-40 between Asheville and Knoxville where a vehicle overturned and caught fire that spread to vegetation, creating a wildfire. The incident and fire created a 12-mile long queue, over six hours of congestion, and low-visibility driving conditions from the smoke. NCDOT requested images around the incident and were able to confirm the crash and wildfire at the scene and enhance TMC situational awareness and response. This demonstrates the value of on-vehicle imagery for special, one-off, and high impact events that are not well captured by existing datasets.

Looking Forward: Integration and Future Applications

The research suggests that integrating on-vehicle camera data into traffic management systems can offer substantial benefits. Future efforts could involve:

• Developing automated tools to streamline the process of searching and combining traffic and weather data with on-vehicle camera images.
• Implementing user interfaces for Traffic Management Center (TMC) operators to view real-time camera feeds, as exemplified in the mockup using Iteris' ClearGuide platform.

Beyond the applications explored in this study, on-vehicle camera data has the potential to support a wider range of traffic management functions, including:

• Real-time monitoring of work zones to track equipment, barriers, and signage.
• Assessing road closures and damage during severe weather events.
• Verifying and responding to Waze alerts for road hazards.
• Monitoring snow accumulation and snow removal efforts.

Collaboration for a Smarter Transportation System

Vizzion focuses on providing real-time traffic insights through their Drives service. This service offers live views and historical footage captured by a network of on-vehicle cameras, giving transportation agencies a highly detailed view of the road.

Iteris specializes in traffic management solutions. Their flagship traffic analytics solution is ClearGuide, which integrates data feeds from various sources, potentially including Vizzion's on-vehicle camera data, to help agencies manage traffic flow and respond to incidents.

The successful collaboration between NCDOT, Vizzion, and Iteris highlights the value of partnerships between government agencies and private companies in developing innovative solutions for traffic management. By leveraging on-vehicle camera technology, transportation agencies can gain valuable real-time insights into traffic conditions, ultimately leading to safer and more efficient roads for everyone.

About the Author:

Jeff Venables is marketing manager, smart mobility applications at Iteris. Connect with Jeff on LinkedIn.

Public link

Please use the above public link if you want to share this noodl on another website.