Chapter 5 Proof-of-Concept Case Studies

This chapter presents case studies on the deployment of Smart Work Zone technologies in three states—Virginia, Pennsylvania, and Iowa—working in collaboration with their respective DOTs. Each state’s engagement resulted in unique applications of the selected Smart Work Zone technologies on projects in their jurisdictions. The individual states were selected for inclusion based on their experience in working with Smart Work Zone technologies, the diversity of operational environment, proximity to VTTI, and willingness to deploy Smart Work Zone technologies in a proof-of-concept testing activity. Experience with Smart Work Zone technologies was assessed through in-person conversations with personnel responsible for work zone technologies within the DOTs. The diversity of the operating environment included factors such as typical roadway geometries, types of work zones being deployed, availability of supporting technologies (connectivity, GPS, etc.), and use of DOT crews versus contractors, etc. Proximity became a factor in our assessment, as VTTI was responsible for making technologies available to each DOT as part of the project and there were practical limitations relative to feasibility of transporting some of the larger equipment such as the ATMA and the supplemental data collection trailer for a short duration deployment.

To avoid redundancy in descriptions, the information on the technologies used and methodologies applied is structured as follows:

• Overview of Work Zone Technologies: A brief description of the individual Smart Work Zone technologies deployed in this project.
• Supplemental Data Collection Systems: An explanation of the additional data collection systems developed and deployed to measure the impact of the Smart Work Zone technologies.
• Performance Metrics: Definition of metrics used to evaluate the performance of the Smart Work Zone technologies.
• Proof-of-Concept Deployments: Detailed accounts of the specific work zone projects, the technologies deployed, and the experience of conducting the deployment.
• Lessons Learned and Suggestions: Insights gained and suggestions for future deployments and technical development.

Overview of Work Zone Technologies

The following sections describe the Smart Work Zone technologies as they were made available for the proof-of-concept deployments. The analysis of technologies in the earlier stages of this project resulted in selected technologies that could be applied independently to support both static and dynamics work zones. The technologies applicable to the static work zones included:

• Work Zone Data Management application (“Work Zone Data Management App”): A tablet-based work zone design, deployment, worker presence, and data management

• software product. This application pictured in Figure 1, facilitates the design and management of work zone data by allowing a user to interactively select an activity area on a map and apply temporary traffic control templates to place signage, channelizing devices, and other necessary elements for ensuring work zone safety. It also produces data that define the detailed locations of the work zone features that can be shared with drivers through a work zone data exchange to help them navigate through work zones more safely. Once designed, the tablet application can be used to navigate to the individual features of the work zone to make sure they are placed in the prescribed locations ensuring that the work zone is set up properly. When work zone status is set to Active, the Work Zone Data Management App begins to publish its data about the work zone to the Virginia DOT’s WZDx and to commercial navigation portals, to then be distributed to consumer applications. This provides drivers with advanced information about the presence and location of work zones, workers, and specific hazards associated with the work zones.

• Smart Work Zone Vests: The Smart Work Zone vests are wearable technology that can localize work zone workers and provide warnings about potential work zone intrusions, alerts about moving into unsafe areas, and warnings when close to moving equipment. For the static work zones included in these proof-of-concept demonstrations, VTTI deployed a Smart Work Zone solution that includes wearable devices, magnetic mount equipment units, and a situation awareness application. The wearable devices pictured in Figure 2 can either be put in the pocket of a reflective Class 3 vest or attached to the accessory mounts on a hard hat. A special magnetic mount unit may also be added to equipment working inside the work zone to provide both location and movement information about them as well. The devices provide highly accurate localization information to a base station that can be mounted on a work vehicle, trailer, or temporary tripod near the activity area. The base station communicates with all devices and determines when to provide individual warnings indicating when the worker is close to, or across a safe area boundary, when a potential work zone incursion is imminent, or when a work vehicle they are close to has just initiated movement. The Smart Work Zone situation awareness application can be used to define the safe areas, manage the geometries around vehicles where warnings will trigger, and parameters for intrusion detection. Example screen shots of this application may be seen in Figure 3. The application shows

• the locations of each individual device and the locations of the safe boundaries. The devices have battery life capable of lasting through a full day shift and can be charged overnight.
• Dynamic Zipper Merge System (DZMS): The DZMS is technology that supports a traffic management strategy used in work zones with lane reductions to optimize traffic flow. Two separate concepts for how to deploy this system were tested during our proof-of-concept activities. If traffic is free flowing, dynamic message signs will instruct drivers to merge early. In one system, sensors upstream from a lane closure detect traffic queues as they develop at the lane closure and when present, drivers are instructed through dynamic roadside signs to use both lanes until the point of merge, then alternate in a “zipper” fashion instead of merging early (see Figure 4). The other system worked in a similar manner but the speeds for the upstream approaching vehicles were derived from probe vehicle telematics data rather than roadside sensors. Several vendors offer these solutions and there are some variances in the wording of the dynamic messaging, etc.

  Most commercially available DZMSs include a means to interact from a remote monitoring system. The remote monitoring system can be used to configure the upstream sensors and dynamic signage, allowing a user to set the thresholds for target speeds where messaging will change, etc. Most vendors of these solutions have integrated their products into a data management process that supports the transmission of data about the work zone to navigation application providers, who then present messages to drivers indicating the location and operating conditions within the work zone using the WZDx data standard. There are a variety of emerging methods to manage this data transfer. The exact methods employed will be discussed in the sections below outlining each work zone experience.

• The remote monitoring systems also receive and archive the sensor data with the historical speed of passing traffic so that it might be used for additional analysis and application development.
• Automated Truck Mounted Attenuator (ATMA): A leader-follower style ATMA was deployed in the dynamic work zone proof-of-concept deployments. The system is capable of operating on limited access freeways and highways in a variety of dynamic and short duration work zones. While still considered experimental and requiring a safety driver to be present, the ATMA is being tested in live work zone operations as pictured below in Figure 5. The vehicle offers the following functionality:

  The ATMA is currently capable of:

  • Operating in both GPS-enabled and GPS-denied environments.
  • Speeds up to 25 mph.
  • Following distances between 50 and 400 feet.
  • Lateral offsets of +/- 12 feet.
  • LiDAR-based automated object detection and avoidance.
  • Temporary stopping (holds and releases).
  • Remote monitoring and fleet operations.

  For this deployment, the system was fitted with a location reporting system and modem that was used to provide real time location information to the WZDx and data dissemination tools.

• Work Zone Data Exchange (WZDx): The WZDx initiative is a collaborative effort led by the U.S. Department of Transportation to improve the accessibility and interoperability of data related to work zones across the United States. This initiative aims to create a standardized format for sharing real-time work zone information among various stakeholders, including transportation agencies, private companies, and the public. WZDx establishes a consistent data feed that provides crucial details about work zone locations, configurations, and impacts on traffic flow. By standardizing this data, the initiative facilitates better communication and coordination, allowing for more efficient route planning and navigation for drivers. It also supports the development of ADAS and autonomous vehicles, which rely heavily on accurate and timely data for roadway events

• such as work zones. The shared data helps mitigate risks associated with work zones such as accidents and delays by informing drivers and traffic management systems about current and upcoming road work.

  At present, several methods exist for making data from Smart Work Zones accessible to drivers. A key data dissemination approach is integrating this data into consumer navigation applications, which can then inform drivers about work zones or reroute them if necessary. Some navigation providers have established partner data portals with technical capabilities to incorporate work zone data into their systems. DOTs are expected to maintain their own WZDx instances and must facilitate this by establishing a service utilizing the latest WZDx data format and ensuring the data is validated and verified. In other data management scenarios, intermediary businesses convert data from a DOT’s work zone management systems into the WZDx format, verify and validate it, and then distribute it to navigation system providers as a service. Additionally, some work zone technology developers have built-in processes for automatically disseminating the data they produce to navigation providers through either of the means listed above.

  For this project and its limited-scale proof-of-concept (POC) deployments, it was deemed impractical to fully integrate our POC technologies into each participating DOT’s mainstream data management processes and systems. However, data from each deployment were shared with consumer navigation applications through multiple channels. The project team established its own WZDx service that provided data in the WZDx format via API and secured approval to post data to a partner portal for a commercial navigation provider. In addition, one technology vendor that was used offered its own data dissemination service to post data to commercial navigation providers through a third-party data aggregation service. The current variety of data dissemination methods present within this industry indicates a lack of a clear, standardized process and architectural standard for managing data exchange with applications that are most likely to make it available to end users. Specific methods used in each scenario will be detailed in the descriptions of the work zone experiences provided in the following sections of this chapter.

Supplemental Data Collection Systems

A goal of this project is to assess the impacts the deployed technologies have on the overall performance and safety of the work zone. Some of the deployed technologies have some built-in ability to provide measurement data while others do not. The project team worked with local DOTs and work zone crews to identify any available data from any systems that were already planned for deployment. However, these sources were not consistent, and it was clear that a supplemental solution would be required.

Smart Work Zone Application Trailer

The trailer is a dual-purpose advanced multi-function observation system designed to monitor operations and safety in work zone environments while gathering data for post-processing analysis. The trailer provides a convenient platform for communications that would otherwise be mounted independently on a work zone vehicle or mobile tripod. The trailer also includes equipment that can be used to monitor traffic flow at the beginning of a lane closure work zone operation using advanced machine vision, radar, and sensor fusion data collection techniques. This solution comprises two primary components:

• Communications technology to support 5G and C-V2X.
• Traffic Monitoring System (TMS) with real-time processing and cloud connectivity.

The diagram below depicts the high-level functionality and architecture of the combined system components installed in the trailer.

Smart Work Zone Base Station

The Base Station provides edge processing and external communications support for Smart Work Zone technologies. It can enhance work zone safety by tracking worker locations using wearable smart vest and smart helmet devices and issuing alerts when workers approach the boundaries of a safety activity area and when they’re near work zone equipment that is about to move. These boundaries can be defined using various methods, including a cone-mounted device, GeoPlotter, or a situation awareness app.

Within a situation awareness app, a work zone supervisor can:

• Monitor the location and status of currently deployed devices (smart vests, smart helmets).
• Track vehicles and machinery equipped with a temporary device within the work zone.
• Configure alert parameters.
• Modify the activity area geofence locations.
• Add notes during deployment.
• Manually broadcast alerts to all smart devices in the work zone.

The system supports cloud communications via 4G/5G modem, which can feed automated worker presence detection, supplementing the work zone design information being sent by the Work Zone Data Management App to the WZDx. The system also supports C-V2X communications where SAE J2735 Personal Safety Messages (PSMs) can be broadcast directly to C-V2X-equipped vehicles approaching the intersection.

Supporting application metrics available from a situation awareness application can include:

• Average number of devices connected by period of day for each device type.
• Utilization metrics
  • Total duration Smart Work Zone technology was active
  • Duration each device was active
  • Number of connections and disconnections by device
• Alerting metrics by individual and collective
  • Number of alerts issued by period of day, alert urgency, and alert type.
• Heat maps showing where alerts were issued

Traffic Monitoring System (TMS)

This component of the system uses cameras and an embedded GPU platform to perform real-time machine vision, radar, and data fusion to track vehicles approaching a position along the roadway. The metrics generated include vehicle counts, average speed, lane selection, vehicle type classification, potential vehicle-vehicle conflict, and potential work zone intrusion information. In the deployments for this project, the trailer was positioned near the final merge point in work zones that included a lane closure to capture behaviors during the zipper merge scenarios. The technical architecture is shown in Figure 6 below.

The TMS is positioned at the base of a work zone where traffic enters the queue. It collects vehicle data that helps users make informed decisions on factors that enhance work zone safety for both workers and drivers. The system uses radar and computer vision to gather data on vehicles passing through a work zone. By employing dual sensors, users benefit from redundancy, which helps reduce errors.

The computer vision component features two cameras aimed at both incoming and outgoing traffic from the system’s location. These cameras feed into a machine learning algorithm that detects vehicles in the images along with lane lines. This setup allows the measurement of vehicle positions, speeds, and lanes of travel. Additionally, the system can manually designate a work zone lane to detect vehicle intrusions.

The TMS is also equipped with two radars, directed at both incoming and outgoing traffic, enabling the measurement of vehicle positions, speeds, and types.

When 5G communications are available, the system offers real-time video streaming for both camera sensors with a processed computer vision layer for remote monitoring and supervision.

The project team worked closely with the work zone crews to find an optimal location for trailer deployment on the roadside based on maximizing field of view for data collection and the safety of both drivers and work crew members.

Supporting application metrics available from the TMS include:

• Traffic Volumes
• Traffic Queue Information
• Vehicle Speeds
• Merging Behavior
• Crashes
• Vehicle Classifications

Telematics Probe Data

In addition to the local data collection trailer, the project team had access to vehicle probe data that are collected by a data aggregation vendor from vehicle telematics systems for some of the work zone proof-of-concept sites.

Proof-of-Concept Deployments

The following sections describe the experiences of conducting the POC deployments in Virginia, Pennsylvania, and Iowa (Figure 7) which represent the culmination of the work performed in this project. The POCs are meant to provide an honest representation of what is required to plan, set up, and execute deployments in operational work zones on public roads while making observations of the deployment experience and how the selected technologies perform. Original project planning would have scheduled our POC deployments to occur late in the 2023 summer and fall road maintenance season. Due to unforeseeable delays in procuring Smart Work Zone equipment and gaining necessary data sharing approvals, our actual deployment window shifted to the 2024 spring maintenance season.

Before the spring 2024 maintenance season, the project team reached out to representatives from each participating DOT to request their involvement. They presented an overview of the project’s goals, details about the POC demonstrations, and the available technologies. They also asked the DOTs to provide a list of scheduled projects within the target timeframe to evaluate their suitability for this project.

Specifically, the POC deployments required the following work zone characteristics:

1. A dynamic or short-duration work zone on a limited access freeway or four-lane highway, or
2. A static work zone on a limited access freeway or four-lane highway that included a lane closure.

The team’s objective was to conduct at least one dynamic and one static POC in each state within the deployment window. Each DOT confirmed they had several suitable projects, some managed by the DOT and others by contractors. Since these projects were already planned and under contract, adding a requirement for contractors to use POC technologies was not possible. Therefore, DOT representatives filtered their project lists to include contractors likely to voluntarily participate in the POC deployments. With a refined list of potential projects and contractors, the project team provided overview presentations to the contractors and ultimately gained their approval to deploy the POC technologies in their work zones.

Virginia Deployments

The Virginia POC deployments included three dynamic and two static work zones operated by a combination of DOT and work zone contractor personnel. The deployments were conducted between April and June of 2024. The Virginia deployments were the first to be executed and the project’s first opportunity to exercise the systems in the field in live work zone operations. A description of each POC is provided in the sections below with the dynamic and static deployments grouped together based on their similarities in type of work zone operation and physical setup.

Dynamic Work Zones in Virginia

The project team conducted multiple dynamic work zones working with VDOT crews to support line re-painting operations in live traffic. For all of these dynamic work zones, the scheduled work projects were matched with the availability of the ATMA system, crew availability dates, and windows of time where weather conditions met the requirements for re-striping operations.

On each day of deployment, a safety briefing was conducted in a pre-determined temporary staging area to emphasize risk mitigation elements built into the vehicle’s operational platform. Since DOT employees were already familiar with standard TMA operations and safety procedures, the focus was on how the extra vehicles (lead and ATMA) would be integrated into standard temporary traffic control (TTC) configuration for a painting operation. A pre-deployment inspection and vehicle walkaround was performed with the crew, highlighting the locations of emergency stop systems, sensors, and any equipment and systems not found on standard TMAs. VDOT instructed the supporting VTTI project team on radio protocols and preferred communication methods for the dynamic work zone operation.

For these deployments, VTTI provided both a leader and ATMA vehicle that would be inserted into the typical TTC-11.1 Moving/Mobile Operations on Limited Access Freeways and Highways with a Single Lane Closure (see Figure 8) between other vehicles in the platoon. The TTC plan was reviewed with the crew and the team confirmed expectations regarding the platoon’s speed (approximately 7 mph), desired following distances (200 ft) between vehicles, and the sequencing of vehicles.

At this stage of development, the safety driver is still required to monitor the ATMA from the driver seat while a sufficient amount of operational data is accumulated to help inform a decision to fully remove the safety driver from the ATMA vehicle. VTTI provided a support driver for the lead vehicle and a safety driver for the ATMA, both experienced with their respective roles in each vehicle’s operations and limitations.

Logistics coordination and approval began with project coordinators working with the VDOT District Traffic Operations Center in Salem, VA, and cascaded down through regional administrators and local supervisors, ultimately reaching individual crew leaders for the project being supported. Several days of operation were finally identified that accommodated all stakeholders. A plan was developed with VDOT work zone safety management to expose the crews to the ATMA technology and supplement their normal safety meetings with ATMA-specific safety information prior to taking the road with the work zone platoon.

The first POC consisted of a dynamic work zone on a four-lane highway near the town of Bonsack, VA (see Figure 9). Once the geographical extent of the day’s striping operations was understood through the briefing, the VTTI team used the Work Zone Data Management App to create a mobile work zone design that covered an outside lane closure between Bonsack and the Bedford County line. This is an approximately 6-mile stretch of four-lane highway with traffic lights and at-grade intersections. The project team updated the Work Zone Data Management App to show the work zone in “Active” status which instantiated the mobile work zone on VTTI’s WZDx as a static definition of the dynamic work zone covering the entire extent of the work zone. After operations were started and the ATMA entered the outer boundary of the mobile work zone, a secondary dynamic definition of the mobile work zone was automatically instantiated in VTTI’s WZDx per the WZDx standard that was updated once per second with the ATMA’s current position.

Once crew briefings were completed, the safety driver positioned the ATMA at its starting position in the platoon and executed calibration and automation initiation sequences, remaining ready to intervene if there were any issues with system startup processes. DOT employees were invited to ride in the passenger seats of either vehicle for observation and to support the generation of feedback on vehicle performance and impact on typical operations methods. A VDOT crew member joined the lead vehicle and monitored the lead vehicle control application and a second crew member rode along in the passenger seat of the ATMA.

After all crew members had taken their stations, the vehicle platoon checked communications and left the staging area for the work zone site. Upon reaching the site the platoon stopped for approximately 1–2 minutes to allow the ATMA and leader to engage automation. Once automation was enaged, the safety driver ceased controlling the ATMA and the painting operation commenced, as pictured in Figure 10 below.

During marking operations, the ATMA demonstrated reliable and accurate performance, maintaining consistent lateral and longitudinal tracking accuracy, including holding static position for extended periods and automatically resuming operations as the paint truck needed to maneuver to address supplemental line painting at turn lanes, etc. However, the ATMA occasionally executed unnecessary holds due to its perception of objects in its path, where no significant static objects were present. These false holds were mostly triggered by vegetation growing close to or over the roadway and by vehicles passing very close to the ATMA in the adjacent lane. The ATMA is designed to stop for such perceived static obstacles but can be quickly cleared to resume once the lead operator confirms there is no hazard. In the case of dynamic targets like passing vehicles, the system holds temporarily until the object is no longer in the travel path and then automatically resumes unless the operator clears it to resume first. After being cleared, the ATMA operates at a pre-defined catch-up speed to rejoin its position within the platoon.

The ATMA successfully completed the shift of operations with the VDOT crew and, while covering 12 miles of striping over 4 hours, represented both inside and outside lane closure configurations. At the conclusion of operations, the Work Zone Data Management App was used to advance the status of the work zone to “Completed,” which removed the work zone definition from the WZDx. There were several technical challenges noted by the VTTI team during the POC 1 testing. The ATMA was not specifically designed to automatically recognize and respond to traffic lights. When the ATMA approached a traffic light and it turned yellow and then to red, the lead vehicle operator was required to initiate a manual hold of the ATMA to keep it from entering the intersection and then release it again when the lights turned green so that it could rejoin the platoon. While response to traffic lights were not a planned component of the ATMA’s operational design domain, the features built into the control system did support operations in the environment and were deemed safe given there was a safety driver in the ATMA. Response to traffic lights is a feature that should be added to the ATMA to have it automatically detect and respond to avoid unsafe behaviors at intersections. The communications link between the leader and ATMA vehicles also failed on occasion and the communications link between the ATMA and the WZDx server that provides its regular position updates did not function properly on the day of deployment. After the POC testing was complete, the VTTI team further attempted to debug the communications issue but were unable to find any direct causes for the communications interruptions.

Similar to the POC 1 deployment, POC 2 consisted of a dynamic work zone on a four-lane highway to conduct stripe re-painting operations in VDOT’s Salem District. The work originated at the Salem City line and continued to the Montgomery County, VA line on US460 (see Figure 11). The logistics coordination and safety briefing were performed with similar District leadership and the same paint crew personnel as in POC 1. For this second deployment, logistics were completed in a more expedited manner due to crew familiarity with the equipment, understanding of goals, understanding of operational limitations, and personnel conducting the research.

A safety briefing was again conducted in the staging area prior to deployment to the working zone. The briefing reminded the crew of the risk mitigation elements built into the vehicle’s operational platform and confirmed where the extra vehicles (lead and ATMA) would be integrated into standard TTC configuration.

The day’s operational plan called for speeds and following distances identical to the prior session. The VTTI team used the Work Zone Data Management App to create a mobile work zone design that covered an outside lane closure between the Salem City line and the Montgomery County, VA line on Route 460. This is an approximately 7-mile stretch of four-lane highway with traffic lights and at-grade intersections.

VTTI provided the support driver for the lead vehicle and a safety driver for the ATMA as in POC 1, both experienced with their respective roles in each vehicles’ operations and limitations. Once crew briefings were completed, the safety driver positioned the ATMA at its starting position in the platoon and executed calibration and initiation sequences. A separate group of DOT employees were invited to ride in the passenger seats of either vehicle for observation and to support the generation of feedback on vehicle performance and impact on typical operations methods. VDOT crew members who had not ridden in the vehicles in the prior session joined the lead vehicle and monitored the lead vehicle control application, and a second crew member rode along in the passenger seat of the ATMA. After all crew members had taken their stations, the vehicle platoon checked communications and left the staging area for the work zone site. Upon reaching the site, the platoon stopped for approximately 1–2 minutes to allow the ATMA and leader to engage automation. Once automation was enaged, the safety driver ceased controlling the ATMA and the painting operation commenced as pictured in Figure 12.

Over the day’s 14 miles of operation, the ATMA had success in many of the same areas as before, including accurate and reliable following maneuvers and response to commanded hold and releases to support the maneuvering of the paint truck. The intermittent communications link failures between the lead and ATMA vehicles were still observed during this deployment. DOT employees provided valuable feedback and suggestions throughout the sessions, identifying several edge cases that are noted in the lessons learned section.

The ATMA support team again conducted troubleshooting to identify the source of the intermittent loss of communications between the leader and follower vehicles and between the ATMA and the WZDx for position information. A power connector for the communication modem was determined to be looser than expected and was thought to be the cause of the intermittent power disconnection to the modem. The connector was replaced, and the communications issue could not be reproduced in further testing. The exchange of position information from the ATMA to the WZDx was found to be inoperable due to software interface incompatibilities. The software issues were also resolved shortly after the POC 2 testing activities were completed. DOT employees participating in POC 2 were generally enthusiastic about the technologies and provided valuable feedback and suggestions throughout the sessions, identifying several edge cases in operations that are noted in the lessons learned section.

The last mobile Virginia POC consisted of a dynamic work zone on a four-lane highway to conduct stripe re-painting operations in VDOT’s Lynchburg District near the City of Lynchburg, VA (see Figure 13). The work began in Campbell County, VA at the intersection of Rt. 24 and US29, continuing up US29 to the Lynchburg City limits. Outreach with the Lynchburg district management was conducted to perform logistical coordination and gain the necessary approvals for deployment.

In this engagement, the safety briefing occurred in the parking lot of the VDOT traffic operations regional office. Like past briefings, the focus was on how the extra vehicles (lead and ATMA) would be integrated into standard TTC configuration for a painting operation. A pre-deployment inspection and vehicle walkaround was performed with the crew, highlighting the locations of emergency stop systems, sensors, and any equipment and systems not found on standard TMAs. VDOT instructed the supporting VTTI project team on radio protocols and

preferred communication methods for the dynamic work zone operation. In this engagement, VDOT crews requested a demonstration of the vehicle’s abilities in the parking lot prior to leaving for the day’s work. The VTTI team accommodated this request, and the ATMA successfully negotiated parked and moving vehicles, as well as infrastructure and other fixed assets through a two-lap loop of the facility.

The TTC plan was reviewed with the crew and the team confirmed expectations regarding the platoon’s speed (approximately 7 mph), desired following distances (150 ft) between vehicles, and the sequencing of vehicles. Once the extent of the day’s striping operations was understood through the briefing, the VTTI team used the Work Zone Data Management App to create a mobile work zone design that covered an inside lane closure between the intersection of Rt 24 and US 29 in Campbell County and the Lynchburg City line. This is an approximately 7.5-mile stretch of four-lane highway with traffic lights and at-grade intersections. The project team updated the Work Zone Data Management App to show the work zone in “Active” status, which instantiated the mobile work zone on VTTI’s WZDx as a static definition of the work zone all along the full distance of the work zone.

After all crew members had taken their stations, the vehicle platoon checked communications and left the staging area for the work zone site. Upon reaching the site the platoon stopped for approximately 1–2 minutes to allow the ATMA and leader to engage automation. Once automation was enaged, the safety driver ceased controlling the ATMA and the painting operation commenced as pictured below.

During marking operations, the ATMA again demonstrated strong performance, maintaining consistent lateral and longitudinal tracking accuracy, including holding static position for extended periods and automatically resuming operations when the paint truck needed to address complex line painting at turn lanes, etc. However, the ATMA experienced some of the similar communications that had plagued it in the past. Initial pairing proved challenging, as did cellular transmissions, known to be a problem in that area.

The ATMA successfully completed the shift of operations with the VDOT crew covering 15 miles of striping over 4 hours, representing both inside and outside lane closure configurations. Poor cellular coverage in the area posed a challenge to communications for team members on site, making it difficult to coordinate vehicle movements accurately. It also prevented the zone from being taken to Active by the on-site personnel directly through the tablet application. As a remedy, off-site personnel contacted and were able to access the work zone definition on the Work Zone Data Management App and update its status to Active.

Operational Challenges Observed

• Communicating operational requirements of the ATMA to crews. While there was an informational and safety meeting prior to executing the POC, there were challenges

• associated with communicating the technical requirements of the ATMA relative to positioning the vehicles in preparation for the engagement of automation. The vehicles needed to be positioned at specific distances apart and within a certain range of headings in order for automation to engage. The crews found these distances difficult to estimate in the field and it took some time to get the positioning correct to allow engagement.
• Maintaining communication links between lead vehicle and the ATMA. A technical issue with maintaining a stable communications link between the lead and ATMA vehicles required the connection to be reset. The loss of communications would disengage automation and cause the crew to have to reset positions and restart automation.
• Maintaining cellular and satellite communications. The terrain and rural environment where operations were conducted was conducive to dropouts in cellular and GPS satellite reception. Loss of cellular communications would temporarily stop updates of the position of the vehicle to the WZDx server. A defect in the software interface also precluded the reliable update of vehicle position to the WZDx for some of the POC testing, but this was resolved during these operations.
• False object detections and holds for vegetation and vehicles passing close by. The ATMA was relatively conservative in its detection and classification of objects in its path of travel, causing it to occasionally stop and hold until released by the lead operator through the tablet control interface. The static vegetation targets required manual verification and release while the dynamic vehicles would clear themselves when the path of travel was found to be clear. This issue was an occasional annoyance and didn’t interrupt the overall flow of operations.
• Higher ATMA operating speeds are limited by forward object detection sensors. Some painting occurred in less densely populated areas, which allowed for faster painting and speeds moving between painting locations. At times, the platoon reached 25–30 mph which was beyond the designed operating speeds for the ATMA vehicle. The vehicle is capable of stable control at the higher speeds but will outdrive the range of its forward object detection and avoidance sensors, which would not allow enough time to detect and respond to any objects there are in its path.

Operational Benefits Observed

• Ease of defining the mobile work zone in the Work Zone Data Management App and making it active on the WZDx. It was relatively simple to define the mobile work zone in the tablet application, and the work zone data was accurate and properly formatted as it was made available through WZDx provided cellular service was available when attempting to do so.
• Accuracy and smoothness of ATMA leader-follower operation. The ATMA followed the leader path and lead operator lateral and longitudinal offset commands very accurately and provided smooth operations. The lead operator was able to set and release hold commands, release the ATMA from its object detection holds, and responded to changes in commanded longitudinal following distance.

• Intuitiveness of ATMA control application controls and user interface. VDOT crewmembers were able to understand the information provided by the tablet application to monitor the response and performance of the ATMA.
• Smooth acceleration and deceleration. The painting operations required frequent holds and releases and was being operated on a hilly terrain with many incline and decline sections. The ATMA vehicle was able to operate smoothly and maintain its commanded headways without abrupt throttle or brake inputs.
• Reliable operation. Apart from the inter-vehicle communications noted earlier, the ATMA was able to operate without loss of control, manual e-stop input, or human override throughout the POC testing session. No significant safety related events or conflicts were observed.

Static Work Zones in Virginia

This POC deployment included a bridge overlay paving project on I-66 in VDOT’s Culpeper District (see Figure 14). The work zone traffic control configuration was a static work zone (TTC-16.2 Outside Lane Closure Operation on a Four-Lane Roadway) on the westbound side of a four-lane limited access freeway (I-66, just prior to Exit 31, The Plains). A DZMS was deployed along 9 miles of the westbound roadway from to Exit 40 (Haymarket) to the merge point with pairs of PCMS located every 3 miles upstream of the beginning of the lane closure.

Prior to deployment, the project team met with the contractor to lay out the proposed plan for deployment in the coming days. Significant communication steps were taken with VDOT and contractor personnel to ensure stakeholders were clear on the participation goals and expectations, including workers wearing and providing feedback on the smart devices, the inclusion of both day and night operational windows, and proper communication channels should any problems occur. The contractor personnel agreed to wear the Smart Work Zone devices and provide feedback to the project team.

The Work Zone Data Management App was used to design the work zone layout and update the work zone status to Active to commence the WZDx data provision (Figure 15). VTTI personnel deployed a Smart Work Zone trailer to both observe driver behaviors traversing the zipper merge and to serve as a power source for the Smart Work Zone base station, which communicates with the alerting devices worn by contractor personnel. The situation awareness app was used to set up a geofence polygon the represented a safe area within the work zone. A total of nine workers in

the work crew wore the wearable devices for the test. To ensure that the trailer’s position would be positioned to capture the merging behavior but not negatively impact any work zone operations, the trailer was deployed a week prior to the work zone becoming active.

The data being captured by the Smart Work Zone trailer were intended to be used to assess the effectiveness of the dynamic messaging being provided by the DZMS. This assessment would largely be dependent on timely data provided from roadway traffic sensors from the vendor systems and their synchronization with the messaging and video data being collected by the Smart Work Zone Technology. Additional coordination was needed with the DZMS regional distributor to establish data transfer requirements and to create an API to support the needed data flows.

During the initial deployment of the Smart Work Zone system, a large number of false alarm warnings were observed by the deployment crew and the workers using the devices. It was determined that the boundaries were defined by too few points for such a large area and additional points would need to be defined to result in a more accurate warning outcomes. The deployment team used the situation awareness application to update the safe area polygon and the performance of the system improved significantly.

The work crews wore the alerting devices for a portion of their shift and were asked to provide feedback on their use and effectiveness. Feedback from workers indicated that the design of the device could be modified to make it easier to insert into pockets of the Class III vests and that power switches could be positioned in a different location to avoide accidentally shutting off power to the device when adding it to their vests. Two comments were made that the overall weight of the device may be an issue after long shifts. In general, the devices were thought to have the potential to provide valuable alterts to the workers.

A second bridge overlay paving project was used for the final Virginia POC deployment, this time on I-81 in VDOT’s Staunton District (see Figure 16). The work zone configuration was a static work zone with an outside lane closure (TTC-16.2 Outside Lane Closure Operation on a Four-Lane Roadway) on the southbound side of a four-lane limited access freeway over Route 600 near Maurertown, VA. A DZMS was deployed prior to the zone for approximately 3 miles of the southbound roadway from Exit 291 (Tom’s Brook) to the merge point with pairs of PCMS every mile. The Work Zone Data Management App was used to design the work zone and to start the WZDx data provision once the work zone was updated to an active status. The Smart Work Zone system was also deployed, and a safe area polygon was established using the graphical user

interface tools. The Smart Work Zone wearable devices were introduced to all crew members and the four workers actively working in the work zone used them for part of their evening shift.

VTTI personnel again deployed a Smart Work Zone trailer in the work zone for data collection and communications support. Stakeholder engagement and advance communication efforts were essential to alignment of all parties’ expectations being met. This zone operated exclusively in the evening hours and required additional support efforts and logistics to ensure support personnel were available and data were collected in an accurate manner. This work zone included a DZMS with roadside sensors, and data from that system was accessed via an API developed for data retrieval.

Observed Operational Challenges

• Optimal alerting for smart devices requires high accuracy when defining the safe area of the work zone. Roadway geometry, particularly in curves, may necessitate additional geofence boundary points to provide a more precise outline of the perimeter of the safe area. Using too few points to define the initial safe area polygon resulted in inaccurate notifications to workers when the devices were initially distributed. Onsite support personnel from VTTI were available to make these adjustments to improve performance in near real time but such adjustments may be too complex or difficult for local crews to make while preparing to conduct their jobs in the work zone.
• Resources designed with multiple uses may have their effectiveness reduced by unforeseen constraints. The Smart Work Zone trailer was designed as a data collection tool (three cameras, two radar devices, remote monitoring capabilities) as well as a power and communications platform for the Smart Work Zone Technology. It may not be possible to position the trailer to optimize its dual role as both a data collection system and a Smart Work Zone base station device. Due to road geometry and other factors, the trailer could not be placed in a single location to optimally perform both of its assigned functions. Collecting traffic behavior approaching the merge point placed it too far away from the zone for reliable communications with the helmet and vest wearable devices and staging it close to the zone for the wearable device support hindered the ability to capture the ideal traffic views.
• The performance of the technology and equipment is affected by its positioning relative to the work zone. The lessons learned on the prior Smart Work Zone regarding the trailer’s inability to serve as a communications hub if parked too far from the zone led to the deployment of a portable base station that allowed for closer proximity to the zone to support the smart technology. The contractor was made aware of the need to locate this equipment near the edge of the zone, and worked with VTTI on a suitable location that did not negatively impact operations. The originally planned location did not suffice to meet the needs of the technology, and an alternative location was eventually found.

• Smart Work Zone technologies are still complex at this level of maturity and requires trained personnel to understand and manage its operation. Due to an inaccurate polygon definition, the devices did not provide alarms reliably and to the full extent of their ability.

Observed Operational Benefits

• Smart Work Zone devices were deemed potentially high value by zone workers. After the safe area geofences were more precisely defined, the devices provided timely and effective alerts to workers that they felt could be valuable. Based on informal worker feedback, the devices could be smaller and potentially lighter but were not thought to be so large or heavy that work would be impeded in their current form. Some ergonomic form improvements were suggested to avoid accidentally operating the power switch.
• Simple, intuitive technology interfaces can be an effective safety tool, while providing accountability and an information management platform. Worker location can be quickly discerned and feedback provided when safe area boundaries are exceeded. The number of alerts can be tracked to the individual worker providing opportunity for accountability if desired.
• Ease of using the Work Zone Data Management App to design work zones and disseminate their data. The Work Zone Data Management App makes it easy to design, setup, and dissiminate data to end users through its integration with the WZDx and other data management tools.
• Flexibility in technology design provides tremendous value. Work Zones are complex and unique entities. No assumptions can be made on equipment location, worker behavior, size/dimensions, or structure. Having the ability to adapt and morph technology and resources that can be adapted to meet the specialized needs of the work is critical to the success of the zone.

Pennsylvania Deployments

Static Work Zones in Pennsylvania

This POC deployment was a bridge overlay and repair project on a four-lane limited access freeway in Pennsylvania’s District 9 near Altoona (see Figure 17). The configuration was a static work zone with an inside lane closure on the southbound side of I-99 just prior to Exit 28 (Ebensburg-Hollidaysburg). Ver-Mac PCMS driven by INRIX data supporting PennDOT algorithms directed the dynamic messaging of the system. The Work Zone Data Management App was utilized to design the zone and publish active statuses for testing WZDx functionality.

Having experienced a recent work zone fatality (April 2024), PennDOT was eager to participate in any research intended to improve negative outcomes. Their TTC leadership helped identify projects that met the minimum POC criteria and assisted in selection and coordination efforts with DOT and contractor project management stakeholders to eventually arrive at the Altoona site.

VTTI personnel transported the Smart Work Zone trailer to the area and positioned it to observe and collect driver behaviors approaching the merge point and through the taper approaching the zone. Once on site, they coordinated with PennDOT and contractor site personnel to verify that the trailer would not have any negative impacts to zone operations.

The data being captured by the Smart Work Zone trailer would ideally validate and gauge the impact and effectiveness of dynamic messaging used in zipper merges. The INRIX traffic flow source data supporting the PennDOT-developed algorithm is housed in a Regional Integrated Transportation Information System (RITIS) managed by the University of Maryland’s Center for Advanced Transportation Technology (CATT) Lab. Additional approvals and access permissions were required for team members before any data analysis or validation could occur.

Data collection efforts occurred over multiple days to capture as many traffic states as possible. The highest-volume queueing was observed between 3:00 p.m. and 4:30 p.m. in the afternoon, with additional queueing at lesser amounts throughout the day.

Observed Operational Challenges

• Communications supporting Smart Work Zone technology can be complex and advance planning for redundancies is needed. The modem supporting remote monitoring of the Smart Work Zone trailer had reached a data threshold that limited its ability to communicate. A backup modem was employed that restored efficient communications for remote support personnel, but consumed focus from the on-site personnel entrusted with configuring critical pieces of the data collection equipment. It also added additional checks for the remainder of the engagement to ensure the integrity of the collected data.
• New technologies and equipment should be discussed as early as possible in the stakeholder engagement process. For this engagement, VTTI had invited the contractor to use the Smart Work Zone alerting devices and provide feedback on the design and performance. After initially agreeing to do so, project leadership requested that their workers refrain from wearing any additional equipment beyond their standard personal protective equipment (PPE). With additional advance notice, future projects would be candidates for device evaluation.

Observed Operational Benefits

• When possible, redundant equipment and backup parts/tools should be taken to any zone to minimize risk and prevent the need to source them near the zone. Having the backup modem and personnel trained in the system setup allowed for continuous monitoring of the zone by subject matter experts in other locations.

Iowa Deployments

Iowa DOT has been at the forefront of work zone safety research and technical advancement for several years, supported in their efforts by Iowa State University and many local contractors. Initial conversations and virtual sessions evaluated many projects, seeking the ideal criteria for the technologies selected for these POCs (static work zone on a multi-lane roadway involving a lane closure and merge). No projects were identified in the needed timeframe that met such criteria. However, other projects were identified that had attractive characteristics suitable for testing the ATMA and Smart Work Zone vests, helmets, and mobile vehicle warning devices, independent of the DZMS. Planning decisions were made to deploy the selected technologies in both static and dynamic work zones in Iowa. With factors such as weather, multi-day vehicle transport from Virginia to Iowa, and project timelines all being considered, a week was identified that provided flexibility to accommodate both the static and dynamic types of deployments. The ATMA and lead vehicle were transported to the Des Moines, IA area, where staging for access to the work zones could occur. Intentional time was blocked for equipment testing and calibration after arrival to ensure performance and technology were not impacted by the transport.

Static Work Zones in Iowa

The Iowa DOT supported the project by helping to arrange for two POC deployments in two static work zones that were operating on two and four-lane roadways with lane closures in each. For the Smart Work Zone vest, helmet, and mobile vehicle device evaluation, coordination efforts initiated in the New and Emerging Technologies group of the Iowa DOT, eventually expanding to the operational work crews and contractor management. Several virtual calls were held to narrow in on available projects that had timelines that would support observational participation by Iowa DOT and other local stakeholder group members. The contractors managing the work zone deployments were willing and especially accommodating to try and give as many options as possible for supporting the POC deployment work.

The first included bridge overlay work on Union County’s Route 34 on a two-lane roadway in a rural setting. The TTC configuration included a temporary traffic signal to manage the alternating flow of traffic to support the lane closure.

The second included a milling operation in a residential neighborhood on Grand Avenue in Ames, IA, involving a four-lane roadway and outside lane closures in both directions.

In both work zone locations, the Work Zone Data Management App was used to design the work zone TTC configurations and published an Active status when appropriate. Internal resources at VTTI verified that the work zone data definitions were posted in the proper format and were available for access from the WZDx data portal. For the Smart Work Zone system deployment, initial safe geofence areas were created in the situation awareness tablet application and the base station that supports device communications was deployed. Subsequent device testing quickly uncovered technical issues that prevented immediate distribution of the devices to contractor personnel. On Route 34 in Union, the Smart Work Zone devices were incorrectly sending alerts while in the safe area of the zone, requiring troubleshooting efforts for debugging and resolution. To provide data and assist in debugging efforts, the devices were instead worn by Iowa DOT participants, who provided feedback on device wearability and ergonomics for approximately 2 hours. Ultimately, the devices were judged to not be alerting with a reliable level of accuracy needed to deploy them with contractor workers in a way that would elicit valuable feedback and a means to gauge their effectiveness. Therefore, the deployment was cancelled. Later follow-up testing found that the algorithms used to calculate a distance measurements to the zone perimeter were negatively affected by specific geometrical attributes of the safe zone polygon, including concavity and the density of available vertices. These issues were corrected both in updates to the system software but also in how the safe area geometries were laid out within the situation awareness app design tool. The system was updated and tested prior to the second deployment on Grand Avenue and improved performance was noted.

The Grand Avenue Smart Work Zone deployment also encountered technology issues but of a different sort. After deploying the base station and defining the safe area polygon geometry, the devices showed improved alerting accuracy as compared to the earlier deployment in Union County. However, there were still times when the reliability of warnings was not to the levels suitable to evaluate effectveness of the alerts. The devices were found to be alerting accurately as designated by the safe area of the zone, but with latency too high to result in an effective alert. When exiting the zone, the devices were consistently providing the initial low-level alerts several feet into the active roadway, most likely too late to provide much value to a worker entering a potentially unsafe area.

The GPS solution that is incorporated into the Smart Work Zone system requires RTK corrections to the standard GPS signal to attain the accuracy required for reliable warning performance. These signals are obtained through a Mobile Network Operator’s distributed network solution that is widely available across the United States. The deployment team verified the system was receiving the supplemental RTK data and that it was properly formatted for use within the GPS receiver. However, for reasons not yet understood, the GPS receiver was intermittently not applying the RTK corrections, causing inaccuracy in the positioning solution and an immediate “jump” in the reported position of the wearable devices. These errors were large enough to cause the device to warn wearers as they were close to or approaching the safe area boundaries. Ultimately, it was decided to not deploy the Smart Work Zone alerting devices on work crew members and the team focused their efforts on making the determinations of what the source of the error was and whether it could be resolved onsite.

Operational Challenges Observed

• Real-world safe area polygon geometry defined in the situation awareness app software revealed a calculation error that resulted in false positive alarms. As referenced above, the layout of safe area geofenced points in a concave structure and with too few vertices resulted in false alerts. To achieve the highest levels of accuracy, the polygon needs additional vertices to properly define the safe area. The distance calculation was updated to accurately handle the concave polygon definitions as well to resolve the reliability issues.
• Unreliable GPS accuracy resulting from unsuccessful application of RTK corrections. Supplemental RTK correction data obtained from commercial supplier networks was received but sometimes not successfully applied to the raw GPS signal, which resulted in degraded position accuracy and false positive alarm generation.
• Technical limitations of Smart Work Zone setup. Initial setup of the “safe area” polygon includes walking a device to various locations and tagging them as vertices of the safe area geometry. The ability to do this automatically with a capture device is limited to when the user is within Wi-Fi range of a single supporting base station device. If the capture device is outside the range of router signal, manual entry of latitudinal and longitudinal points is required, introducing possibility for error and adding to the setup time.
• Weather impact to Smart Work Zone infrastructure. The communications base station was intentionally designed with a level of weatherproofing sufficient for most expected conditions in work zones. However, it is dependent on an external 120 V power source and a Power over Ethernet (PoE) injector in the base station unit. The external power source, as well as supporting laptops and tablets, are not adequately protected in rainy conditions and require additional protection to support Smart Work Zone operations.

Operational Benefits Observed

• Ease of instantiating work zone data flow with the Work Zone Data Management App. The tool was found to make work zone design simple. The tablet-based application makes it easy to begin data transfer to a WZDx outlet by simply updating a status to an Active state.

Dynamic Work Zones in Iowa

Several opportunities to deploy dynamic work zones in Iowa were explored with the Iowa DOT stakeholders and several were identified that would be appropriate for POC deployment. The results of those deployments are reported below. While their operations were conducted on separate roadways, the configuration of the dynamic work zone and results of operations were similar, so they are being reported together in the following section.

The mobile POC consisted of a dynamic work zone on a four-lane highway to conduct stripe repainting operations in Boone County, which spans Iowa DOT District 1 between Ames and Ogden. It involved multiple vehicles and an inside lane closure as operations were performed. A lead vehicle and ATMA were integrated into the Iowa DOT platoon for the duration of the operation (see Figure 21).

After successful testing of the ATMA the prior afternoon, the VTTI team met Iowa DOT crews at their operations and maintenance facility in Ames for the morning safety briefing. The day’s planned painting operations west of Ames on Route 30 were reviewed with the operations team. The project team addressed immediate questions about the truck’s ability to function as a standard TMA if the automation were to malfunction, as well as whether the cushion met existing standards. Iowa DOT currently employs units with TMA intrusion alarm systems and crews are versed in their operation, so familiarity existed with additional components designed for worker safety. As with all previous deployments, an inspection and vehicle walkaround was performed, highlighting the locations of emergency stop systems, sensors, and any non-standard TMA equipment. Iowa DOT personnel instructed the project team on radio protocols and preferred communication methods for the dynamic work zone operation. A paint supervisor joined the ATMA safety driver for the initial deployment. All other crew members were needed for vehicle operation, so no additional Iowa DOT personnel rode in the lead vehicle.

Iowa’s standard TTC for pavement marking operations is TC433, which calls for multiple TMAs and support vehicles (see Figure 22). The TTC plan was reviewed with the crew, and the team confirmed expectations regarding the platoon’s speed (approximately 10 mph), desired following distances (250 ft) between vehicles, and sequencing of vehicles. The VTTI team used the Work

Zone Data Management App to create the static definition for the mobile work zone covering the approximately 17-mile stretch of four-lane highway with traffic lights and at-grade intersections. Once operations commenced, the zone was set to “Active” status, instantiating the mobile work zone on an WZDx. Once the ATMA entered the geographic outer limits of the mobile work zone, a secondary definition of the mobile work zone is created within WZDx, which provides more frequent updates of the ATMA’s current position within the larger work zone definition.

During marking operations, the ATMA demonstrated excellent performance, maintaining significantly improved communications from past deployments. It held a longitudinal position with expected responsiveness as in past engagements, including holding static position for extended periods and automatically resuming operations when the paint truck needed to address complex line painting at turn lanes, etc. The ATMA successfully executed manual hold maneuvers at a four-way stop on the morning route. However, after being cleared of the hold, the ATMA exhibited a slow response to a vehicle that crossed into its path of travel and a manual takeover of vehicle control was executed by the ATMA safety driver. The issue has been noted and tagged in the data collection system and an investigation into the root cause of the incident and corrective development actions are in process.

Aside from the incident mentioned above, the remainder of the painting operations were completed without incident. After westbound operations terminated in Ogden, the Work Zone Data Management App was updated to reflect an eastbound mobile zone with inside lane closure for the route back towards Ames. No issues were experienced during the return trip, including additional opportunities for manual hold and resumption of operations at four-way stops. Eastbound work was eventually interrupted by intermittent rain and terminated prior to reaching the original starting point.

The weather cleared, providing an opportunity for additional work that same afternoon closer to the crew’s home facility in Des Moines. After a brief assembly at the Ames facility for coordination, a different paint supervisor swapped into the ATMA passenger position and the platoon proceeded south, queueing at the north Des Moines facility. The Work Zone Data Management App was used to develop and manage the status of dynamic work zones that covered the planned areas of operation. Painting operations then resumed on Route 415 at 54^th street, proceeding outbound for approximately 3 miles using a mobile left lane closure. TC433 was employed with the same sequence of vehicles used in the morning session, but with reduced spacing due to the higher-density traffic environment. The ATMA performed exceptionally well, performing four lead vehicle-initiated holds/releases for red lights, as well as dynamic adjustments of following distances (100 ft to 150 ft and back to 100 ft), and holding consistent communications throughout the session. No significant technical issues were experienced in the afternoon session.

Operational Challenges Observed

• Object detection reliability at intersections. A vehicle conflict at a four-way stop in Boone required the safety driver to intervene to control the ATMA. The ATMA also executed an object detection hold for (seemingly) a vehicle in close lateral proximity. The operator approved the ATMA to continue and it functioned as anticipated. The ATMA held a second time for an unknown reason and the hold was manually overridden and the ATMA continued on its path.

Operational Benefits Observed

• Accuracy and smoothness of leader-follower operation. The ATMA followed the leader path and lead operator dynamic longitudinal commands very accurately and provided smooth operations. The lead operator was able to set and release hold commands, release the ATMA from its object detection holds, and responded to changes in commanded longitudinal following distance.
• Intuitiveness of ATMA control application controls and user interface. Iowa DOT crewmembers were able to understand the information provided by the tablet application to monitor the response and performance of the ATMA. They suggested possible improvements to the ATMA system, specifically video-game style controls (steering wheel and accelerator/brake pedals) to support occasional manual operation and additional cameras positioned on the ATMA for greater visibility and awareness of ATMA status. In addition to the existing forward and rear-facing cameras, they proposed side cameras with views perpendicular to the ATMA’s lane of travel.
• ATMA Metrics. After collecting data from POC1, POC2 and POC8, the team analyzed the ATMA deployment and generated Table 12 with the most important metrics from all three POCs, including morning and afternoon deployments. This table includes a rate for Hold and Disengagement events which were calculated considering the deployment duration.

Table 12. ATMA Metrics

POC Lessons Learned and Suggestions

The following summarizes some lessons learned and suggestions that can be gleaned from these POC deployment experiences.

• Any technology designed for use by work zone personnel on-site should be simple, intuitive, and easily implemented. Ensuring work zone technology is simple and robust is vital for roadside workers due to safety and efficiency concerns. Simple, easy-to-implement technology reduces the learning curve and minimizes distractions, allowing workers to focus on their tasks, thus reducing accident risks. Robust equipment withstands harsh roadside conditions, ensuring continuous operation and minimizing downtime. With limited access to technical support, straightforward and durable technology empowers workers to resolve issues independently, maintaining workflow efficiency. This approach fosters confidence and compliance with safety protocols, enhancing overall work zone safety and efficiency as well, benefiting both workers and the traveling public. Simplicity of operation is a key element in determination of product maturity in this space.
• Many of today’s emerging work zone safety technologies rely on reliable external systems, such as communications links and GPS signals. Work zones can be found anywhere there is roadway or transportation infrastructure. Often, these areas are supported by nearby communication infrastructure that facilitates data transfer, including critical application data. However, communication access is not always ubiquitous. Even in densely populated urban areas, there can be gaps in coverage or limitations in available bandwidth. As a result, technologies dependent on external communications may become unusable or ineffective without reliable access. To address this, redundancies in communication systems and alternative methods should be integrated to ensure functionality in less-than-ideal conditions at work zone sites. The same considerations apply to GPS signal reception, which can be significantly affected by obstructions, weather phenomena, and solar radiation activity. Users must be aware of these limitations and understand how they might impact the deployment of technology in various environments.
• Early and consistent communication between all stakeholders cannot be underestimated and should be a primary consideration when setting up and executing work zones. As work zone technology evolves, safety initiatives are often spearheaded by the technology specialty areas within DOTs. These departments typically have limited interaction with the operations crews handling daily tasks. Each group, including subcontractors, operates under distinct management layers. It is crucial to avoid assuming that these often-separate groups share the same priorities or fully understand the work’s goals and objectives. Therefore, it is essential to develop a robust communication plan that incorporates flexibility and redundancy to reduce risks. While there is a risk of overcommunicating with groups not constantly engaged with technology, it is important to use discretion and tailor communication levels to meet the specific needs of each stakeholder group. Effective and timely communication helps prevent the repositioning of equipment, the redundant and inefficient use of resources, and issues with deliverables and execution, ensuring that proper expectations are set and maintained.
• Work Zones can have similarities regarding objectives, type, and duration, but each zone is unique and new technologies should be adaptable to meet specific needs. While work zones may share common objectives, types, and durations, each work zone is inherently unique, presenting its own set of challenges and requirements. These zones are designed to ensure the safety and efficiency of both workers and motorists, often involving similar tasks such as traffic control, lane closures, and road maintenance. However, specific circumstances, such as location, traffic patterns, and environmental conditions, can vary significantly from one work zone to another. Therefore, it is crucial for new technologies implemented in these zones to be highly adaptable and customizable. This flexibility

• ensures that the unique needs of each work zone are met effectively, promoting safety, minimizing disruptions, and enhancing overall project efficiency. By leveraging adaptable technologies, work zones can better address their specific demands and improve their operational outcomes.
• Execution of work zones can be fluid, and technologies must be adaptable and flexible. Managing work zones requires a fluid approach, utilizing adaptable and flexible technologies. Despite thorough and extended planning efforts, DOT projects often face re-prioritization and dynamic changes right up until the start of work. Therefore, maintaining flexibility and readiness for last-minute adjustments is crucial. This flexibility applies to all resources, including personnel, equipment, and schedules. It’s advisable to avoid using equipment designed for a single purpose. Instead, whenever possible, utilize multi-use resources and adapt them as needs evolve. Work zones often have space and access limitations that can be worsened by specialized tools or equipment. These tasks could often be handled by existing resources on site.
• The generation of high quality, accurate work zone data requires integration of technology into a DOT’s work zone business processes. The generation of high-quality, accurate work zone data from technologies necessitates the integration of systems into a DOT’s work zone business processes. Automation and digital tools help in collecting precise data, which enhances decision-making and safety, but governance of data and authority to publish must be addressed. Implementing such technologies ensures that data is consistently updated and accurately reflects current conditions, leading to improved traffic management and reduced congestion. This technological integration is vital for optimizing work zone operations and enhancing overall efficiency and safety on the roads.