Alternative Technologies for Mitigating the Risk of Injuries and Deaths in Work Zones: Conduct of Research (2022)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

20 CHAPTER 3: Findings and Applications This chapter presents the results and analysis from each research step described in Chapter 2 along with discussions of the practical application of the findings. Specifically, the chapter describes the results of the literature review, documentation and evaluation of WZITs, online survey, and technology case studies. The chapter also presents and describes the outputs of the research for use in practice (i.e., DSS and guide). 3.1 Literature Review As described in Chapter 2, a comprehensive literature review of online, publicly available resources was conducted to gather pertinent information related to the WZITs and to support the development of the industry survey questionnaire, case study interviews, and guide. The results of the literature review as described below. 3.1.1 Types and Availability of WZITs The literature review revealed a long list of technologies for work zone traffic control that are presently commercially available or under development, many of which are specifically designed, or may be used to, prevent and mitigate work zone intrusions. Technologies that are intended, or can be used, for work zone intrusions can be grouped into the seven broad categories based on technology type and intended application. Table 3.1 shows the seven technology categories along with 28 examples of specific technologies within the categories. Each WZIT is classified according to its primary technology.

21 Table 3.1. WZITs WZIT Type/ Application Example Technologies Motorist Vehicle Systems 1. Autonomous vehicles 2. Connected vehicles (e.g., V2V) Work Zone Construction Equipment Systems 3. Autonomous equipment 4. Connected equipment 5. Automated equipment with truck-mounted attenuators 6. Mobile barrier 7. Automated flagger with intrusion alert device UAS 8. UAS for signage 9. UAS for monitoring Intrusion Detection and Alert Systems 10. Intrusion alert system with equipment-mounted sensors (e.g., AWARE) 11. Intrusion alert system with cone/barrel-mounted sensors (e.g., SonoBlaster) 12. Intrusion alert system with networked cone/barrel system with sensors (e.g., Intellicone) 13. Intrusion alert system with pneumatic tubes (e.g., Worker Alert System) 14. Intrusion alert system with Bluetooth 15. Intrusion detection with computer vision and ranging (e.g., SmartCone) 16. Intrusion alert system with RFID 17. Queue warning system with networked cone/barrel system (e.g., iCone System) Enhanced Signage and Enforcement 18. Work zone warning with dynamic message system (DMS) 19. Speed warning with DMS 20. Speed enforcement with radar 21. Speed enforcement with photo radar Wearables 22. Wearable lighting (e.g., Halo Light) 23. Head-mounted augmented reality (AR) display 24. Smart watches/bracelets 25. Smart vest for workers Other Technologies 26. Worker safety through simulations 27. Robot-controlled automated flagger stations 28. Device-free localization (DFL) 3.1.2 Technology Capabilities and Features The suite of available WZITs that were identified have different capabilities, features, and qualities. A variety of means are used for sensing and monitoring, including: radar, pneumatic tubes, radio signals (i.e., RFID), infrared light, cameras (still and/or video), light detection and ranging (lidar), GPS, ultra-wideband (UWB), and BLE. Similarly, the types of outputs generated by the technologies to alert workers and to provide information about the site and intrusion varies

22 amongst technologies. Visual, sound, vibration, and electronic data are common types of outputs generated. The capabilities and features of each category of technologies are described in further detail below. 3.1.2.1 Motorist Vehicle Systems Motorist vehicle systems refers to those technologies in which the safety application is, in part, embedded in the passing vehicles. The technologies utilized help to prevent vehicle intrusion by improving real-time driver awareness through the use of a warning message or a notification to the drivers as they enter a work zone. An alert message is also commonly sent to the workers who may be affected if an intrusion occurs. The technologies in this category are autonomous vehicles (AVs) and connected vehicles (CVs), which are described further below: 1. Autonomous vehicles (AV): AVs are self-driven vehicles that do not require assistance from a driver to navigate the roadway. AVs are powered through technologies such as Global Navigation Satellite System (GNSS) systems, sensory systems, and radar systems, and have built-in Bluetooth, gyroscope, and mechanical systems. As roadway work zone construction is a dynamic environment which does not have common pre-defined activities, the challenge lies in the movement of AVs through the work zone. AVs are generally operated on artificial intelligence systems and machine learning systems where the environment is continuously monitored and complex actions are executed simultaneously to run the vehicle. AVs have been used in work zones through special pavement markings. A study by Singh and Islam (2020), aimed to use special pavement markings to assist with AV navigation in work zones. The pavement markings are constantly interpreted by the AV systems and allow a smoother driving experience for the vehicle. The work zone pavement markings also enable the AV to be driven under the speed limit set within the work zones (Singh and Islam, 2020). 2. Connected vehicles (CV): CVs are generally operated with a dedicated short-range communications (DSRC) system to which the communication from the user to the vehicle is through a fixed wavelength to provide in-vehicle information for the drivers. CVs prove to be effective in improving traffic flow and to reduce congestion on freeways due to work zones. In-vehicle information includes, but is not limited to, providing warnings of upcoming variable speed limits, warnings of queues ahead, lane closures warnings, and vehicle-to-vehicle messages. CVs can provide real-time feedback through messages displayed on variable message signs communicated via road sensors (UMass Transportation Center, 2019). The applications extended to improving driver response, where it was observed that 20 professional drivers were seen to reduce vehicle speed when a notification regarding adverse weather conditions given to the driver, and a smooth braking response was recorded around work zones within a driving simulator environment (Raddaoui et al., 2020). 3.1.2.2 Work Zone Construction Equipment Systems Highway construction employees utilize various construction equipment to complete tasks and activities in work zone settings. Recent developments have been made in utilizing technologies

23 within the construction equipment to alert operators of the equipment and workers on the roadway surrounding the construction equipment when an intrusion occurs. This type of technology is a relatively new class of work zone intrusion systems, and is termed “work zone construction equipment systems.” This category includes: autonomous equipment, connected equipment, autonomous equipment with truck-mounted attenuators, mobile barriers, and automated flagger with intrusion alert device. 3. Autonomous Equipment: Autonomous equipment is generally classified as similar to AVs and automated mechanical machinery, and referred to as smart equipment. Autonomous equipment typically has an operator, but also has sensor and radar systems to automatically collect information about the surrounding environment to determine the next feasible step to execute the operation (Manufacturers, A. of E., 2021). The application is further extended to construction equipment that is present in the work zone to detect the work zone boundary and not allow operation outside the boundary. The terrain changes are constantly updated on the 3D map in a computer at a remote location through the sensory systems. This capability enables the equipment user to decide necessary actions to finish the task (Welles, 2020). 4. Connected Equipment: Connected equipment is similar to CVs, except that the DSRC systems are used in the equipment in multiple devices connected to a single network and deliver real-time status and location due to being connected to a set of traffic control and roadway infrastructure standards. An example of connected equipment is the iCone, which can be integrated into various pieces of work zone safety equipment and transforms the features of the equipment into a type of connected equipment. iCone can be installed into arrow boards, iPin, and flagging baton devices that enable relaying the information of upcoming highway construction activity to the vehicle users upstream of the work zone (iCone Products, 2020). This application was found to be of a particular interest by developing a hazard detection algorithm tested in the Virginia Connected Corridor (VCC), where a web-based situational awareness tool assesses the hazard detection between the workers, CVs, and connected equipment (Han, 2019). 5. Autonomous Truck Mounted Attenuator (ATMA): Highway maintenance operations are usually mobile work zones in nature, and commonly involve some type of striping and sweeping activities. In order to alert distracted drivers who are on a path to crash into the maintenance vehicles in the work zones, automated equipment is attached to the truck- mounted attenuators (TMAs) as a modification, such as a Directional Audio System that raises an alarm when a vehicle intrusion is detected and, hence, provides a safety measure to the workers (Brown et al., 2015). Unmanned TMAs consist of an unmanned follower vehicle “driven” from a leader vehicle. GPS navigation of the follower vehicle enhances safety in the work zone and eliminates any threat to the driver who is absent from the follower vehicle (Kratos, n.d.). 6. Mobile barrier: A mobile barrier is a mobile mechanical system that acts as a barrier between the passing vehicles and the work zone for the purpose of protecting the workers from vehicle intrusions. Mobile barriers are mobile, easy to set up and remove in and around short-term work zones during the peak traffic hours, and minimize roadway traffic

24 congestion (Mobile Barriers, n.d.). Mobile barriers are useful during the daytime and can be used during the nighttime as some of the barriers have a lighting system. Mobile barriers commonly have a programmable message sign/board attached, and are NCHRP 350 TL-3 crash compliant-barriers. A TMA may or may not be attached to provide a cushion for the impact of vehicles from the rear end of the truck (Hallowell et al., 2010). As reported in other studies, the mobile barrier was found to have state of the art safety barrier functions against crosswise vehicle intrusions. The mobile barriers were found to absorb energy from a crash through crushing on impact and the TMA (Kamga and Washington, 2009). 7. Automated Flagger Assistance Device (AFAD): An AFAD is an automated version of the traditional flagger operation which can be either operated manually or remotely by a flagger. An AFAD is used to control traffic into the work zone. The AFAD is positioned on the roadway shoulder to reduce the exposure to the traffic flow. There are two types of AFADs available, one type is remote-controlled with a slow-stop sign, and the second type utilizes a red-yellow lens to indicate when motorists can enter the work zone. Both types of AFADs are used to alternate the right-of-way through a work zone (Trout et al., 2013). The entire operation entails a mechanically operated temporary traffic control device setup around work zones and is operated automatically with the programmed setup for the duration of stops and movements. An automated flagger is usually used for short-term or intermediate-term lane or road closures. All AFAD applications abide by the specific standards set forth in MUTCD Section 6E.04, and are in accordance and satisfy the crash worthiness standards based on the device weight. Detailed specifications for slow/stop AFADs are available in MUTCD Section 6E.05, and MUTCD Section 6E.06 for the red/yellow lens AFADs (ATSSA, 2020a). 3.1.2.3 UASs UASs are a completely mechanicalized aircraft system consisting of three system components: (1) an autonomous or human-operated control system which is usually on the ground or a ship, but may be on another airborne platform; (2) an unmanned aerial vehicle (UAV); and (3) a command and control (C2) system - sometimes referred to as a communication, command, and control (C3) system to communicate with the remote user. The highway work zone safety applications using UASs are getting attention and a few of the examples are provided below. 8./9. UAS for Signage and Monitoring: A UAS consists of a UAV along with a camera system that provides real-time traffic monitoring with high temporal and spatial resolution which helps to dynamically manage the work zone by interacting real-time with changeable message boards, driver mobile phones, and traffic management center and law enforcement agencies (Malveaux et al., 2020). UAS drone data has shown to be a potential application in roadside asset inventory (including signs and culverts), and as built documentation of the work zone project, classification in the right-of-way, and the video and images are useful for communication with the public (Hubbard and Hubbard, 2020). In another set of experiments, the UAS proved to successfully decrease the mean speed by 2 km/h during the UAS-based drone temporary traffic control (TTC) operations (Barlow et al., 2020).

25 3.1.2.4 Intrusion Detection and Alert Systems An older class of technologies that has been researched and developed, and is commercially available is intrusion detection and alert systems, or simply abbreviated as IAS. The IAS concept is based on installing and extracting the technologies primarily on TTC devices. Some new IASs are standalone technologies that can be used in and around work zones to provide safety to the workers and/or the motorists who intrude in a work zone. Examples of IASs include: intrusion alert system with equipment-mounted sensors (e.g., AWARE), intrusion alert system with cone/barrel-mounted sensors (e.g., SonoBlaster), intrusion alert system with networked cone/barrel system with sensors (e.g., Intellicone), intrusion alert system with pneumatic tubes (e.g., Worker Alert System), intrusion alert system with Bluetooth, Intrusion detection with computer vision and ranging (e.g., SmartCone), intrusion alert system with RFID, and queue warning system with networked cone/barrel system (e.g., iCone System). 10. Intrusion Alert Systems (IAS) with Equipment-mounted Sensors (e.g., AWARE): This category of IASs contains equipment-mounted sensors that utilize a single technology or multiple technologies such as radar (scanned radar), high-precision differential GPS system, accelerometers, gyroscopes, and magnetometers, for position and orientation sensing to detect vehicle intrusions. The radar is spread out up to 200 feet in the covered area in a funnel or a semicircular shape covering the work zone lane and adjacent through lane. The AWARE system is an example of this type of IAS, where the AWARE uses target threat detection and tracking technology to assess the vehicle speed, location, and probable trajectory. Although the AWARE system is not available commercially, testing has been conducted to evaluate the warning lights and audible alarm in the lane closure, lane closure in right and left curves alignment, along with flagging and tangent alignment operations. This system can be utilized in a long-term work zone duration (Theiss et al., 2017) 11. IAS with Cone/Barrel-mounted Sensors (e.g., Sonoblaster): These sensors are embedded in the IAS devices and are generally impact-activated work zone intrusion devices. The intrusion devices are mounted on a cone/barrel and activated when a vehicle hits the cone/barrel during an intrusion impact. An example is a SonoBlaster, which is comprised of a CO2 cartridge and an alarm unit. Research studies provide advantages and disadvantages of this type of device. False negatives have been reported due to the accumulation of ice between the CO2 cartridge and the firing pin, which produces only a short burst of sound after the first use (Sanni, 2019). In another study, a cumulative response rate through three field tests showed 92% and 85% worker response rates at 50- and 100-foot distances from the IAS, respectively (Gambatese et al., 2017). The use of a SonoBlaster was suggested in work zones around low-speed roads. However, while there were some concerns raised as the cones with the SonoBlaster units cannot be stored, the setup and removal take more time than anticipated, some SonoBlaster units sound the alarm during set up due to poor supervision, and the arming of the unit is difficult (Wang et al., 2013). 12. Intrusion Alert Systems with Network Cone/Barrel System with Sensors: This type of IAS typically consists of radio-based intrusion systems. These systems use radio waves to

26 facilitate two-way communication between sensors and the alarm unit (Awolusi and Marks, 2019). The Intellicone system is an example of such a system. Intellicone is a motion sensitive cone lamp and web enabled device equipped with general packet radio service/global system for mobile (GPRS/GSM) communications and GPS sensors. The Intellicone is installed on a channelizer device, such as a cone, using a bolt or other means to secure it in place (Sanni, 2019). The main components are: a portable site alarm, traffic management unity, integrated lamps and sensors with a Sentry motion sensor. This system houses various options such as communications between system components in the field and with the central command, site alarm and a traffic management unit (Trans Canada Traffic, Inc., n.d.). 13. IAS with Pneumatic Tubes (e.g., Worker Alert System): Pneumatic tube-based systems use pneumatic tubes that are connected to a transmitter. When run over by a vehicle, the tube activates its alert mechanisms, such as an alarm siren and/or strobe light, alerting workers and drivers. The tubes are generally place parallel to the road shoulder and nearby the advance warning area of the work zone. The alert mechanisms are activated when a vehicle passes over the pneumatic tube (Sanni, 2019). The Worker Alert System (WAS) is an auditory and visual alarm system designed and developed to be wirelessly triggered when a vehicle crosses over a positioned pneumatic hose in a work zone. The pneumatic hose has a pressurized sensor that triggers the portable case alarm up to a 1000 ft when a vehicle crosses over the hose. The auditory and visual alarms are activated for about 6 seconds (Astro Optics, n.d.). 14. IAS with Bluetooth: This system contains a network of Bluetooth zones through a chip that are integrated with a single command and processing center to act as an intrusion alert system. Several devices utilize the BLE based proximity sensing and alert system, which uses adaptive signal processing (ASP) methods to communicate between the Bluetooth zones. This application not only extends to detect close proximity sensing, but the automated collection of information also enables sharing e-notices to drivers and diverting traffic flow (Park et al., 2016). In another series of experiments, the Bluetooth technology proved to be a reliable and appropriate alarm system with minor performance differences. The alert display function was positively validated at a real-time construction site and performance results showed a positive response by the participating equipment operators and on-site work zone workers (Marks et al., 2017). The field results indicated that the WorkzoneAlert app is reliable for detecting BLE beacon signals at an average distance of 127m away from traffic signs or portable radar speed signs, and successfully announces the corresponding message associated with each BLE beacon (Lioa, 2019). 15. Intrusion Detection with Computer Vision and Ranging (e.g., SmartCone): The concept of computer vision and ranging in a work zone intrusion alert device is executed with the use of cameras, the IoT concept, and LiDAR to provide real-time feedback in either auditory or visual modes. Computer vision and ranging generally utilizes artificial intelligence and machine learning to continuously learn about the environment, keep track of the objects within the range and learn about their physical attributes, assess the danger or threat, and provide real-time feedback to the user. SmartCone, for example, houses these rich features that enable the early warning system through detection by the high-level

27 technology which is safe and easy to use. The warnings are sent out early on and provides mission-critical feedback to people in the area and any personnel that need to be notified (SmartCone, n.d.). 16. IAS with RFID: RFID is a technology that utilizes radio signals to communicate between devices and is present in popular applications in work zones. RFID technology generally consists of three components: a tag/smart label, a reader, and an antenna. The tag embedded in a particular asset transmits radio signals to the reader. Each reader has a processor that converts radio signals into a digital form so the data can be uploaded into a database (EZ Office Inventory, 2019). In one study of an RFID system, a driver smart advisory system (DSAS) effectively helped reduce vehicle speed before a work zone when vehicles approached the first and second temporary traffic signs, but no reduction was found at the third traffic sign (Qiao et al., 2014). The researchers developed a system of devices that can successfully detect the presence of a forklift, dozer, and excavator in the work zone. Extended testing in various environments showed promising results using auditory, visual, and vibrational alarms (Allread et al., 2009). 17. Queue Warning Systems (QWS) with Networked Cone/Barrel System (iCone System): In this system, traffic safety control equipment is mounted on a cone or within a cone and series of cones to form a network to relay information to the workers and drivers to provide safety, improve travel times, and aid real-time decisions as the vehicles approach the work zone. While a queue warning system is not designed as a work zone intrusion device, its use in connection with other technologies could perhaps provide warning of a potential intrusion. A similar idea is seen in the iCone system where a network of cones utilizes Wi-Fi and collects information about the roadway conditions, and transmits the information onto a cloud server for inferential understanding. iCone marks virtually anything that is on the roadway that causes a vehicle to alter speed and/or direction (iCone Products, 2020). 3.1.2.5 Enhanced Signage and Enforcement Innovative technologies which utilize visual attraction to notify the vehicle users of upcoming work zones and local law enforcement agencies around the work zones to lower the speed have been developed over the past few decades. These types of technologies, which typically show a positive influence on the vehicle users, include: work zone warning with DMS, speed warning with DMS, speed enforcement with radar, and speed enforcement with photo radar. 18./19. Dynamic Message Signs (DMSs) for Work Zone Warning and Speed Warning: Dynamic message signs have been employed in highway work zones as an innovative TTC device in the United States for many years. Their development and use prove to be effective at decreasing vehicle mean speed before the work zone through numerous changes, such as an animated message displayed instead of a text-based message, radar attached to the message sign that notifies the drivers of their speed, and a change in the rate a message flashes per minute to influence driver attention and behavior (ATSSA, 2020b). Test results indicate that sign placement and content have no significant impact on speed reduction and compliance (Bai et al., 2015). Sign

28 frame refresh rate and the word “worker” were found to have a significant effect on a driver’s initial speed and speed reduction (Rahman et al., 2017). In another study, the deployment of a portable changeable message sign (PCMS) helped to reduce the speed of passenger cars and trucks, and the researchers recommended that the deployment location should be determined through the experience of traffic engineers (Bai et al., 2015). In one of the latest studies conducted, the behavioral response of drivers to compromised or hacked DMS boards in work zones was recorded (Ermagun et al., 2020). The study findings showed a mixed impact on vehicle speed and driver distraction. This result signifies concern for safety and efficiency of future transportation works (Ermagun et al., 2020). 20./21. Speed Enforcement with Radar and Photo Radar: An automated speed enforcement (ASE) system is an enforcement technique with one or more motor vehicle sensors producing recorded images of motor vehicles traveling at speeds above a defined threshold. The speed limits are set by police officers and speeds are monitored through radar technology. Images captured by ASE systems are processed and reviewed in an office environment. All citations are reviewed and certified by a police officer. Violation notices are then mailed to the registered owner of the violating vehicle. Often, ASE systems are referred to as speed cameras (Pennsylvania DOT, 2021). 3.1.2.6 Wearables The use of personal wearable devices by workers helps the workers to visualize the work area and be visible to drivers in the event of a vehicle intrusion during nighttime work. This technology class has recently been getting attention as it involves advanced technological methods which have potential benefits to safeguard the workers in the work zone. A list of technologies in this category are as follows: wearable lighting (e.g., Halo Light), head-mounted AR display, smart watches/bracelets, and smart vests. 22. Wearable Lighting (e.g., Halo Light™): Wearable lighting consists of wearable devices worn by the workers during nighttime work to illuminate the work area and alert others of their presence. Standard are available that prescribe the minimum illumination required by wearable lights to qualify to use as a wearable safety device. Generally, a highly reflective tape is worn around the vest along with a strobe light wrapped around the hardhat powered by a battery attached to the worker’s body. The lights flash to indicate to the driver in the vehicle of the presence of the worker on the roadway. Products such as a Halo Light™ are worn on the top of the hardhat to ensure sufficient lighting is visible to the driver. A recent research study found that the combination of both wearable personal lighting and roadway lighting helped the workers to be more visible to the motorist and the equipment operators in the work zone (Jafarnejad et al., 2018). 23. Head-Mounted AR Display: AR is a technological-tool that superimposes virtual visual data on real-life objects. AR applications are used in construction project scheduling, progress tracking, worker training, safety management, time and cost management, and quality and defects management (Ahmed, 2018). AR has been proven to be an information

29 browser through interaction with the visual data and is seen in hybrid interactive systems with BIM models, computer vision, and cloud computing concepts and methods (Chen and Xue, 2020). AR technology is found to be applicable in constructability reviews through real-time work zone site assessment of the safety risks, and ultimately helps to secure safer work zones (Mallela et al., 2020). 24. Smart Watches/Bracelets: Smart watches and bracelets are electronic devices warn on the wrist which have the capability to connect to a wireless system through Bluetooth and Wi-Fi systems. The devices enable workers to read messages and get alerts/warnings through flashing lights or vibration. Development of the technologies for work zones use, especially to alert workers of intrusions, is still in progress. 25. Smart Vests: Smart vests are safety vests with a warning and communications system integrated into the vest. Smart vests may utilize Bluetooth, GPS, radio sensors, lighting systems, and other technologies to enable the worker to receive alerts and have sufficient time to avoid a vehicle collision or intrusion. Smart vests are powered up through portable batteries attached to the vest. The vests may also contain sensors to obtain information about the vital statistics of the worker such as heart rate, level of fatigue, and strength level. Other information gained are the location of the worker within the work zone of the construction site. Work zone applications are being developed. One example is the “InZoneAlert” vest powered by radio sensors that provides a 5 to 6 second warning before a collision occurs. Simultaneously, the vest also warns the driver of a timely collision which would happen within a few seconds (Grose, 2015). The SmartVest is another example, which has yet to be investigated in the connect work zone environments, specifically wherein the communication channels established in the SmartVest are investigated during vehicle intrusion threats (Mollenhauer at al., 2019). 3.1.2.7 Other Technologies Other new developments in technology or the technological concepts that are in progress or yet to be developed are discussed in this section. These include enhancing worker safety through simulations and DFL. The technologies are in the spectrum of initial stages to research and development, as is the robot-controlled automated flagger stations technology. 26. Worker safety through simulations: Various simulation software programs are available and utilized as decision support tools (Abdelmegid et al., 2020). Simulations can be used as an acceptable solution when the construction process is difficult to describe or the relationship of the issues in the project is difficult to process (Nikakhtar et al., 2011). Crash simulation tests are carried out under guidelines of test Level 2 of the NCHRP report 350. The study concluded that both the weak and rigid soil simulations were within acceptable limits and the intrusion alarm systems can be deployed on the soil (Burkett, 2009). Work Zone Driver Model 1.0 (FHWA v1.0) was upgraded to FHWA v2.0 with enhanced capabilities by interfacing with PTV Vissim (microscopic multi-modal traffic flow simulation). The FHWA 2.0 software was tested in Springfield, MA. Test results showed improved performance as it predicted queue lengths, queue locations and travel speeds more accurately than FHWA v1.0 (Berthaume et al., 2020).

30 27. Robot-controlled automated flagger stations: Robot-controlled automated flagger stations are self-driven autonomous and robotic flagger systems that have the ability to direct and monitor traffic flow and provide real-time alerts to workers and motorists ensuring safety to the workers in the roadside construction work zone. An industry survey was conducted to assess the technology and features required in the initial stages to start out the project (Bushe et al., 2020). The study found that innovative tools such as artificial intelligent cameras, RFID, LiDAR, and transportation systems are desired to track the vehicles, recognize the speed, and alert workers of an eminent threat of a vehicle intrusion (Bushe et al., 2020). 28. DFL: DFL encompasses extensive use of wireless local access networks (WLANs) and mobile devices for the interest of worker localization in an environment through device- based localization systems which use GPS and RFID based technologies (Pirzada et al., 2014). In one study, a framework was built to analyze the workers in and around the construction site with a high-risk profile. Multi-path effects were reinforced along with hands free device localization to track workers (Edirisinghe et al., 2014). The concept is yet to be extended and explored in the capacity of work zone vehicle intrusion surroundings. 3.1.3 Technology Functionality and Uses Preventing and mitigating work zone intrusions requires that multiple functions be performed, either by a technology or a human. The general functions that are required are as follows: • Perimeter monitoring • Location monitoring (e.g., where the intrusion through the perimeter has occurred) • Issue warning to workers and/or drivers • Intrusion prevention (e.g., positive barrier) • Intrusion mitigation (e.g., lessen the impact after an intrusion occurs) • Connectivity to work zone infrastructure and personnel However, every WZIT typically does not perform all of the functions. Some technologies only perform one function, while others can perform multiple functions. In addition, the functions that can be performed may be performed at different points in time relative to an intrusion. For example, some technologies are able to warn workers before an intrusion occurs. Other technologies protect workers after an intrusion occurs. Understanding the capabilities of each WZIT, i.e., what and when the technology performs its designed function(s), is an important part of selecting a WZIT for implementation. The technology capabilities can then be mapped to the needed functionalities associated with intrusions to optimize technology implementation. The research team classified these functionalities chronologically relative to the occurrence of an intrusion as follows: • Before intrusion: o Warn workers of potential intrusion o Warn drivers of potential intrusion o Provide a barricade/barrier to prevent intrusion

31 • During intrusion: o Detect intrusion o Alert workers of intrusion o Alert drivers of intrusion • After intrusion: o Protect workers from intrusion o Protect drivers from intrusion 3.1.4 Technology Applications The available WZITs also vary in terms of when, where, and how the technology is applied. In some cases, a technology may be placed on the equipment used for the work operation, while other technologies may be located on the roadway surface. The means in which a technology is applied, and the types of work operations, roadway conditions, and work zone environments for which it is particularly suited, can impact its feasibility for use in practice. Table 3.2 provides application characteristics that are commonly considered when selecting WZITs for implementation.

32 Table 3.2. Applications of WZITs Application Description Application Options Location of technology in work zone Advance Warning Area Transition Area Activity Area – Buffer Space Activity Area – Work Space Termination Area Placement of technology On construction equipment On roadway surface – lane On roadway surface – shoulder/median On roadway feature On person Aerial Other Timing of technology deployment Project Planning Project Design Construction - Planning Construction - Mobilization Construction – Performance of the Work Construction - Demobilization Type of work zone/ Duration of technology deployment Long-term (>3 days) Intermediate (1 – 3 days of daytime work, or >1 hour of nighttime work) Short-term (>1 hour of daytime work) Short-duration (<1 hour) Mobile Type of work Construction Maintenance Roadway/environmental condition Dry Wet Icy Time of day Nighttime Daytime Type of work activity Paving Sweeping Drainage cleaning Lamp/signal maintenance/replacement Bridge maintenance Road widening Barrier repair/installation Striping Utility installation/maintenance Other

33 Table 3.2. Applications of WZITs (continued) Application Description Application Options Location of work activity Off road On shoulder On travel lane Other Vehicle speed in work zone Less than 35 mph 36 – 45 mph Above 45 mph Roadway location Urban Suburban Rural Roadway type Local street Arterial roadway Highway Expressway/freeway Number of lanes Two lanes (one lane each direction) Four lanes (two lanes each direction) More than four lanes (more than two lanes each direction) Roadway separation Undivided Divided 3.1.5 Technology Readiness for Implementation In addition to the technology capabilities/features, functionality/uses, and applications, technology adoption depends on other factors both associated with the technology and external to the technology. Initial cost and availability of the technology are important factors considered to assess its feasibility for widespread implementation. The maturity of the technology is important as well. Technology maturity is commonly assessed through a TRA. TRA protocols have been developed that provide a relative TRL for a technology. The TRL can then be used to compare one technology to another in terms of its readiness for adoption and implementation. The FHWA Technology Readiness Level Guidebook (FHWA 2017) is an example of a TRA process that has been developed and is used in the transportation industry. TRA processes are also available in other industries, such as that TRA process developed by NASA. Starting with the FHWA guide, and integrating aspects of other TRA processes tailored specifically to construction operations, the researchers developed the technology readiness levels shown in Table 3.3.

34 Table 3.3. TRLs used in TRA Phase TRL Description and Assessment Questions/Requirements Basic Research 1 Basic research: Initial scientific research has been conducted. Principles are qualitatively postulated and observed. Focus is on new discovery rather than applications. i. Do basic scientific principles support the concept? ii. Has the technology development methodology or approach been developed? 2 Application formulated: Initial practical applications are identified. Potential of material or process to solve a problem, satisfy a need, or find application is confirmed. i. Are potential system applications identified? ii. Are system components and the user interface at least partly described? iii. Do preliminary analyses or experiments confirm that the application might meet the user need? 3 Critical function or proof of concept established: Applied research advances and early-stage development begins. Studies and laboratory measurements validate analytical predictions of separate elements of the technology i. Are system performance metrics established? ii. Is system feasibility fully established? iii. Do experiments or modeling and simulation validate performance predictions of system capability? iv. Does the technology address a need or introduce an innovation in the field of transportation? Applied Research 4 Components validated in laboratory environment: Design, development and lab testing of components/processes. Results provide evidence that performance targets may be attainable based on projected or modeled systems. i. Are end-user requirements documented? ii. Does a plausible draft integration plan exist, and is component compatibility demonstrated? iii. Were individual components successfully tested in a laboratory environment (a fully controlled test environment where a limited number of critical functions are tested)? 5 Laboratory testing of integrated/semi-integrated system: System component and/or process validation is achieved in a relevant environment. i. Are external and internal system interfaces documented? ii. Are target and minimum operational requirements developed? iii. Is component integration demonstrated in a laboratory environment (i.e., fully controlled setting)?

35 Table 3.3. TRLs used in TRA (continued) Phase TRL Description and Assessment Questions/Requirements Development 6 Prototype demonstrated in relevant environment: System/process prototype demonstration in an operational environment (beta prototype system level). i. Is the operational environment (i.e., user community, physical environment, and input data characteristics, as appropriate) fully known? ii. Was the prototype tested in a realistic and relevant environment outside the laboratory? iii. Does the prototype satisfy all operational requirements when confronted with realistic problems? 7 Prototype demonstrated in operational environment: System/process prototype demonstration in an operational environment (integrated pilot system level). i. Are available components representative of production components? ii. Is the fully integrated prototype demonstrated in an operational environment (i.e., real-world conditions, including the user community)? iii. Are all interfaces tested individually under stressed and anomalous conditions? 8 Technology proven in operational environment: Actual system/process completed and qualified through test and demonstration (pre-commercial demonstration). i. Are all system components form-, fit-, and function-compatible with each other and with the operational environment? ii. Is the technology proven in an operational environment (i.e., meets target performance measures)? iii. Was a rigorous test and evaluation process completed successfully? iv. Does the technology meet its stated purpose and functionality as designed? Implementation 9 Technology refined and adopted: Actual system proven through successful operations in operating environment, and ready for full commercial deployment. i. Is the technology deployed in a work zone? ii. Is information about the technology disseminated to the user community? iii. Is the technology adopted by the user community? As described in the FHWA guide, the TRL scale can be used as a guide to structure discussions about the state of development (or maturity) of a single technology. To achieve a specific TRL, the technology must meet all of the requirements within that level and prior levels. Using the TRL values and descriptions shown in Table 3.3, the researchers conducted assessments of the 28 identified WZITs. This assessment was verified by the subject matter experts. Table 3.4 shows the TRLs for each of the technologies. It should be noted that TRLs are subjective and should be used for general guidance only. Discussion and assessment should be performed using a group of experts knowledgeable about the technology, industry, and current practice.

36 Table 3.4. TRLs for identified WZITs WZIT Type/ Application Technology TRL Before Intrusion During Intrusion** After Intrusion Warning for Workers Warning for Drivers Barrier Detect Intrusion Alert Workers Alert Drivers Protect Workers Protect Drivers Motorist Vehicle Systems 1. AVs 9 5.5 8.5 2. CVs (e.g., V2V) 7 6.5 Work Zone Construction Equipment Systems 3. Autonomous equipment 3.5 2 3.5 3 4. Connected equipment 3.5 3.5 3.5 5. Automated equipment with TMAs 8.5 6.5 6.5 8.5 6. Mobile barrier 8.5 9 7. Automated flagger with intrusion alert device 4 7.5 5.5 5 7 UAS 8. UAS for signage 6.5 9. UAS for monitoring 2.5 3.5 1.5 1.5 Intrusion Detection and Alert Systems 10. Intrusion alert system with equipment-mounted sensors (e.g., AWARE) 8 8 8.5 8.5 8.5 11. Intrusion alert system with cone/barrel-mounted sensors (e.g., SonoBlaster) 8.5 8.5 5 12. Intrusion alert system with networked cone/barrel system with sensors (e.g., Intellicone) 8.5 8.5 5 13. Intrusion alert system with pneumatic tubes (e.g., WAS) 8.5 8.5 5 14. Intrusion alert system with Bluetooth 7 5.5 7.5 6.5 15. Intrusion detection with computer vision and ranging (e.g., SmartCone) 7.5 8 16. Intrusion alert system with RFID 2 2 17. Queue warning system with networked cone/barrel system (e.g., iCone System) 7

37 WZIT Type/ Application Technology TRL Before Intrusion During Intrusion** After Intrusion Warning for Workers Warning for Drivers Barrier Detect Intrusion Alert Workers Alert Drivers Protect Workers Protect Drivers Enhanced Signage and Enforcement 18. Work zone warning with DMS 9 19. Speed warning with DMS 9 20. Speed enforcement with radar 9 21. Speed enforcement with photo radar 9 Wearables 22. Wearable lighting (e.g., Halo Light) 8.5 8.5 23. Head-mounted AR display 2 24. Smart watches/bracelets 7.5 25. Smart vest for workers 3.5 Other Technologies 26. Worker safety through simulations 27. Robot-controlled automated flagger stations 2.5 2.5 6.5 2 2 28. DFL 1.5 1.5 * Technology readiness levels are provided only where the technology provides functionality. Blank cells indicate that the technology does not have ability to perform that function. ** Detection is required for alerting, and detection is immaterial if an alert is not provided. Therefore, in cases where a technology can only perform one of these functions during an intrusion, the technology must be used in conjunction with another technology that provides the other functionality.

38 3.1.6 Technology Adoption Adoption of a technology within an organization or on a project is a decision made by those with authority over the organization/project. When deciding whether to adopt a new technology, multiple factors are often considered. Adoption factors are the qualities, features, and conditions deemed important in the assessment of a technology when considering whether to acquire and implement the technology. Research shows that there are a variety of different adoption factors that are considered. The adoption factors that have been identified in prior safety technology research (Nnaji et al. 2019) and are specific to WZITs (Eseonu et al. 2018) are presented in Table 3.5. As shown in the table, the factors can be grouped into three different categories: technology- related, organization-related, and external-related, which is consistent with the technology– organization–environment (TOE) framework (Baker 2012). The adoption factors listed in the table were incorporated into the industry survey as described below. Targeted strategies implemented by the adopting organization can facilitate technology adoption and implementation. Such strategies enable successful implementation for those technologies deemed worthy of adoption. Prior research has identified effective strategies for promoting innovation adoption and implementation (Darko et al. 2018, Nnaji and Karakhan 2020). Those strategies that are particularly applicable to WZITs, and which were included in the industry survey, are shown in Table 3.6. For convenience, the strategies are organized into four categories based on the type of strategy and when/how it is implemented: standards and regulatory-based, awareness-based, financial incentive-based, and operations-based. It should be noted that adoption of a technology requires awareness and understanding of the available technologies and their potential uses. Education and training are needed for those employees who will champion the technology adoption process within a state DOT. Therefore, DOTs need to provide and facilitate opportunities for awareness and learning about the technologies. Additional resources outside of specific projects may need to be allocated for such a technology awareness/training program.

39 Table 3.5. Safety technology adoption factors for WZITs Factor Category Adoption Factor Technology-related Level of resistance to environmental impact (rain, heat, etc.) Having the required features/technical attributes (warning alert sources, power source, etc.) Level of technical support required Ability of warning to alert driver/motorist (motorist comprehension of warning signal) Ability of warning to alert worker (worker comprehension of warning signal) Coverage distance Frequency of false alarms Number of warning alert sources (visual, audio, haptic, etc.) Durability (reusable, battery life, etc.) Level of complexity (ease of deployment, movement, retrieval, maintenance, storage, etc.) Ability to limit worker exposure to traffic Organization-related Level of technical support available Level of training required Potential level of resistance from employees (responsibility, privacy, etc.) How quick users will be able to influence colleagues Potential cost savings from using the technology Level of compatibility with current processes (impact on traffic control setup, etc.) Competitive advantage derived from using the technology Top management degree of involvement (championing the adoption or not) Cost of labor, equipment, maintenance Organization culture (receptive to change or not) External-related Industry-level change requires technology adoption Government policy and regulation Direct competitors adopt similar technology Client demand

40 Table 3.6. Strategies for effective adoption and implementation of WZITs Strategy Category Strategy Standard and Regulatory-based Mandatory work zone safety technology policies and regulations Better enforcement of work zone traffic control policies after the WZITs have been implemented Availability of competent and proactive WZIT promotion teams and local authorities More WZIT adoption advocacy by the Federal Highway Authority Development of WZIT certification program Awareness-based Creation of public highway safety awareness through workshops, seminars, and conferences Publicity through media (e.g., print media, radio, television, and internet) WZIT-related educational and training programs for vendors, contractors, DOTs, and policy makers Availability to collect information on cost and benefits of WZIT Financial Incentive-based Financial incentives for WZIT adoption Low-cost loans and subsidies from government and financial institutions to develop and pilot-test emerging WZIT Strengthened WZIT research and development through additional research investment Acknowledging and rewarding WZIT adopters publicly Operations-based WZIT assessment programs Availability of institutional framework for effective WZIT implementation Support from executive management within the firm/organization Inclusion of specific technology requirements in contracts Employee involvement in implementation decision- making 3.1.7 Technology Operation Once a technology is adopted, consideration must be given to its continued operation and implementation. Consideration is given to operational costs, storage requirements, maintenance needs, and external power requirements. Ease and complexity of use are also considerations, and should be considered during deployment, retrieval, and transportation of the technology. Its use must also be considered relative to existing traffic control configurations and practices. Importantly, the technology must be effective to justify its continued use. Effectiveness can be

41 measured in terms of whether the technology efficiently and accurately performs the intended functions and provides a positive return on investment. 3.2 Survey of Practice The research study included an industry-wide survey to collect information about the use of WZITs in current practice and the perspectives from industry personnel about the effectiveness of the WZITs and common strategies for their adoption and implementation. Survey development was informed by the initial background work, described above, related to the types of WZITs available, the WZIT capabilities/features, functionality/use, applications, and readiness, and the WZIT adoption factors and strategies identified. 3.2.1 Survey Questionnaire The researchers developed a set of survey questions based on the information collected regarding WZITs and the objectives of the study. The survey included questions in the following areas: respondent personal and organization demographics, work zone safety technology use timeline, and perspectives on the effectiveness of WZITs and factors influencing successful implementation of WZITs. The survey questions were submitted to a panel of subject matter experts (SMEs) for their review and input. The survey questions were revised accordingly based on the input received prior to its distribution. A copy of the survey questionnaire is provided in the Appendix. As shown above, Tables 3.1 and 3.4 provide a list of 28 WZITs and representative TRL values for each technology, respectively. Based on the TRLs identified for each technology, and to create an efficient survey questionnaire that would not require a long time to complete (and therefore help increase response rate), the researchers reduced the list of technologies addressed in the questionnaire. In some cases, technologies that received low TRL values were dropped from the list, and in other cases technologies were combined into one representative technology due to their similarity. As a result of this reduction, the list was reduced to the following 15 technologies, which were included in the survey questionnaire: 1. Connected and AVs 2. Automated equipment with TMAs 3. Mobile barrier 4. Automated flagger with intrusion alert 5. UAS for signage 6. Intrusion alert system with equipment-mounted sensor (e.g., AWARE) 7. Intrusion alert system with cone/barrel-mounted sensor (e.g., SonoBlaster) 8. Intrusion alert system with networked cone/barrel-mounted sensor (e.g., Intellicone) 9. Intrusion alert system with pneumatic tube sensor (e.g., WAS) 10. Intrusion detection with Bluetooth 11. Intrusion detection with computer vision and ranging (e.g., SmartCone) 12. Queue warning system with networked cone/barrel sensor (e.g., iCone System) 13. Dynamic/changeable message sign and speed enforcement 14. Wearable lighting (e.g., Halo Light™) 15. Smart watches/bracelets

42 3.2.2 Survey Distribution The survey questions were uploaded to an online survey distribution tool (Qualtrics) for distribution. Prior to its distribution, the survey documentation and protocol were submitted to the Institutional Review Board (IRB) at Oregon State University for review and approval for research involving human subjects. The researchers then sent invitations, via email, to the contact list developed for the survey sample. Emails were sent to at least one person in each state DOT, plus at least one person in an industry organization (GC, subcontractor, WZIT manufacturer/vendor, or other) in 48 states across the U.S. A link to the online survey was included in the email. The emails requested their participation in the survey and provided background information about the study. Multiple follow-up emails were sent in the following weeks to encourage participation and increase the number of responses. 3.2.3 Survey Results and Analysis A total of 134 responses to the survey were received. Responses were deemed sufficiently complete if the respondent completed all of the questions at least through Question #7: “How long has your organization used WZITs on roadway construction projects?” Additional incomplete responses, straight liners, and speeders were eliminated from the data set prior to the analysis. This data management process ensures that only quality responses are included in the analysis. As a result of applying these criteria and efforts, a total of 101 responses were deemed sufficiently complete to include in the data set. The following sections provide descriptions of the respondent pool based on their demographic information (location, experience, title, etc.), followed by summaries of their responses related to WZIT use in practice and their perspectives of WZIT performance. 3.2.4 Survey Results: Respondent Demographics Table 3.7 and Figure 3.1 present the distribution of responses that were received based on the respondent’s organization. The table summarizes the total number of responses received (n = 134) along with the sufficiently complete responses (n = 101). In terms of states represented, sufficiently complete responses were received from DOTs in 40 different states (80% of states), and from industry organizations in 10 different states (20% of states). Multiple responses were received from many states. In addition, in some cases, survey responses were received from both DOT and industry personnel in the same state. As a result, the total number of states identified in Table 3.7 may be more or less than 50.

43 Table 3.7. Survey responses by type of respondent organization Organization All Responses Sufficiently Complete Responses Total # of States* Total # of States* State DOTs 64 41 62 40 Industry (contractors, subcontractors, WZIT manufacturers/vendors, other) 43 10 39 10 Not specified 27 Not available 0 0 Total 134 101 * Some states may be common between DOT and industry responses; hence the total may be more or less than 50 states. Figure 3.1. Distribution of responses by type of respondent organization As depicted in Figure 3.1, for those responses that were sufficiently complete, the majority came from employees of state DOTs, which amounted to approximately 61% of the responses. Approximately 34% of the sufficiently complete responses were submitted by personnel working in construction general contractor or subcontractor firms. Figure 3.2 shows the distribution of responses across the U.S. based on state. Of the 101 sufficiently complete responses, 98 of the respondents included the state in which they work. Responses (both DOT and industry responses) were received from all states except the following eight states: IL, NE, NJ, NM, ND, OK, RI, and VT. More than one response came from 21 states, with 12 and 14 responses coming from Alabama and Florida, respectively. Multiple responses

44 were received from each region in the U.S. as identified by the Bureau of Labor Statistics: West (13), Mountain-Plains (12), Midwest (11), Southwest (2), Southeast (43), Mid-Atlantic (7), New York-New Jersey (2), and New England (8) (BLS 2014). The geographic spread of participants in the study helps to ensure that the results presented are relevant across the U.S. * States with responses, but too small to show on map: Connecticut (4), Delaware (1), Hawaii (1), Maryland (1), and Massachusetts (2). Figure 3.2. Distribution of responses by state (n = 98) Figure 3.3 shows that approximately 21% of the participants are project managers, while 15% are traffic control designers or engineers. In addition, 13% of the responses were received from individuals in safety specific roles (safety officers, safety engineers, safety managers, etc.). In all, responses were received from a diverse population, thereby ensuring the study accounts for multiple perspectives.

45 Figure 3.3. Distribution of responses by respondent title/position The responses came from individuals with many years of experience. Sixty-nine of the 101 (66.4%) sufficiently complete responses came from individuals with more than 20 years of experience in the transportation/construction industry. In addition, those with 11-20 years of experience amounted to approximately 21% of respondents, while approximately 9% of respondents have 6-10 years of experience. The many years of work experience within the survey participants provides a high level of confidence that the input received is informed by a substantial wealth of knowledge about the industry. Most of the survey participants (63 of the 101 responses, or 62%) work in organizations that employ over 1,000 workers, followed by 16% who work in organizations with 251-500 employees, 8% who work in organizations with 501-1,000 employees, and 8% who work in organizations with 101-250 employees. Most state DOTs employ over 1,000 workers, which is a key reason behind the skew in responses based on organization size. The survey responses recorded within Qualtrics were downloaded for analysis using statistical analysis software. Provided below is a summary of the responses related to WZIT use in practice. Only those responses that were deemed to be sufficiently complete (n = 101) are included in the summaries provided. 3.2.5 Survey Results: Respondent Familiarity with WZITs Using a binary option in the survey (i.e., Yes or No), participants were asked to indicate if they are familiar with the 15 WZITs for one or more functions (protecting drivers, warning workers, etc.). As shown in Figure 3.4, most participants have some level of familiarity with wearable lights, 21.05% 6.32% 14.74% 12.63% 15.89% 12.63% 6.54% 4.21% 8.42% Percent of Responses based on Respondent Title/Position (n = 101) Project Manager Project Engineer Traffic Control (Work Zone Designer/Engineer) Safety Officer Others Chief/Assistant Chief (design, operation, maintenance, etc.) Construction Manager Superintendent Company owner

46 DMS, and mobile barriers (88% of responders selected Yes). Most participants were not familiar with UASs for signage (83%), intrusion detection with Bluetooth (79%), intrusion alert with pneumatic tubes (74%), and intrusion detection with vision and ranging (74%). This finding is somewhat consistent with previous WZIT research (Gambatese et al. 2017). Figure 3.4. Respondent familiarity with WZITs When comparing responses received from DOT employees (n = 61) and non-DOT participants (n = 39), results indicate that DOT employees are typically more familiar with WZIT than their counterparts (see Table 3.8). The difference in familiarity level between DOT employees and non- DOT workers was above 20% for seven out of the 15 WZITs assessed, which suggests that contractors are notably less familiar with WZITs. However, contractors are slightly more familiar with the UAS for signage, wearable lights, and smart bracelets.

47 Table 3.8. Respondent familiarity with WZITs: DOT vs. non-DOT Work Zone Intrusion Technology Familiarity with WZIT (% of respondents) Yes No DOT Non-DOT DOT Non- DOT Connected and autonomous vehicles 69 54 31 46 Automated equipment with truck-mounted attenuator 82 74 18 26 Mobile barrier 93 77 7 23 Automated flagger with intrusion alert 79 49 21 51 UAS for signage 16 18 84 82 Intrusion alert system with equipment-mounted sensor (e.g., AWARE) 69 38 31 62 Intrusion alert system with cone/barrel-mounted sensor (e.g., Sonoblaster) 61 33 39 67 Intrusion alert system with networked cone/barrel-mounted sensor (e.g., Intellicone) 48 26 52 74 Intrusion alert system with pneumatic tube sensor (e.g., WAS) 34 13 66 87 Intrusion detection with Bluetooth 23 18 77 82 Intrusion detection with computer vision and ranging (e.g., SmartCone) 31 18 69 82 Queue warning system with networked cone/barrel sensor (e.g., iCone System) 62 15 38 85 Dynamic/changeable message sign and speed enforcement 97 74 3 26 Wearable lighting (e.g., Halo Light) 87 90 13 10 Smartwatches/bracelets 46 49 54 51 3.2.6 Survey Results: Duration of Use of WZITs Survey participants were asked how long their organizations have used WZITs on roadway projects. Figure 3.5 reveals that over 55% of the participants’ organizations do not currently use WZITs on roadway projects, while only 6% of the organizations have used WZIT for over 10 years. Interestingly, research and application of WZITs dates to the 1980’s according to Nnaji et al. (2020). Responses from DOT and non-DOT participants were almost identical.

48 Figure 3.5. Duration of WZIT use To determine the current rate of WZIT implementation, the survey questionnaire asked participants to indicate if their organizations have: used a WZIT in the past but discontinued its use; are currently using WZITs on their projects; currently not using WZIT but plan to use in the near future; or have no plans to use WZITs. Interestingly, only two technologies are currently being used by more than 50% of the participants’ organizations: DMS (63.9% of respondents) and wearable lighting (51.2% of respondents). Furthermore, over 50% of participants indicated that their organizations do not plan to use intrusion detection with Bluetooth (57.1%), UAS for signage (53.3%), intrusion alert system with cone/barrel-mounted sensor (53.1%), intrusion detection with computer vision and ranging (52.0%), and intrusion alert system with pneumatic tube sensor (50.0%). Figure 3.6 shows the results with respect to each technology evaluated. 8.0% 4.0% 4.0% 6.0% 22.0% 56.0% Percent of Responses based on Duration of WZIT Use (n = 101) 1 to 3 years 4 to 5 years 6 to 10 years More than 10 years I dont know Not currently used

49 Figure 3.6. State of WZIT use

50 Table 3.9 provides a comparison between DOT and non-DOT responses with respect to the extent to which WZITs are used by the organization. While over 43% of DOT participants indicate that their organization is currently using or plans to use connected and AVs, 80% of non-DOT participants indicated that their organization does not plan to use this technology in a work zone. To ensure successful implementation of technologies such as connected and AVs, DOTs should work closely with contractors to ensure buy-in. Contractor responses indicate that they are currently using the following WZITs at a higher rate than the sampled DOTs: automated equipment with TMA; mobile barrier; dynamic/changeable message sign and speed enforcement; and wearable lighting.

51 Table 3.9. State of WZIT use: DOT vs. non-DOT Work Zone Intrusion Technology Extent of Use of WZIT (% of respondents) Used previously but discontinued Currently using Not using but plan to use soon No plan to use I don't know DOT non-DOT DOT non- DOT DOT non- DOT DOT non- DOT DOT non- DOT Connected and AVs 0.0 0.0 7.3 0.0 36.6 0.0 29.3 80.0 26.8 20.0 Automated equipment with TMA 0 0.0 20.8 33.3 25.0 7.4 25 40.7 29.2 18.5 Mobile barrier 3.7 7.1 30.0 50.0 13.0 10.7 40.7 21.4 13.0 10.7 Automated flagger with intrusion alert 4.4 11.8 24.4 17.7 24.4 11.8 24.4 41.2 22.2 17.7 UAS for signage 0.0 0.0 22.2 0.0 0.0 33.3 55.6 50.0 22.2 16.7 Intrusion alert system with equipment-mounted sensor 4.9 14.3 2.4 0.0 26.8 14.3 39.0 50.0 26.8 21.4 Intrusion alert system with cone/barrel-mounted sensor 2.7 0.0 2.7 8.3 13.5 0.0 51.4 58.3 29.7 33.3 Intrusion alert system with networked cone/barrel- mounted sensor 0.0 0.0 0.0 0.0 11.1 0.0 44.4 60.0 44.4 40.0 Intrusion alert system with pneumatic tube sensor 0.0 0.0 23.8 0.0 4.8 0.0 42.9 80.0 28.6 20.0 Intrusion detection with Bluetooth 0.0 0.0 7.1 0.0 14.3 0.0 57.1 57.1 21.4 42.9 Intrusion detection with computer vision and ranging 0.0 0.0 5.6 0.0 11.1 0.0 44.4 71.43 38.9 28.6 Queue warning system with networked cone/barrel sensor 2.78 0.0 38.9 20.0 16.7 20.0 22.2 40.0 19.4 20.0 Dynamic/changeable message sign and speed enforcement 1.8 3.7 62.5 66.7 12.5 7.4 12.5 18.5 10.7 3.7 Wearable lighting (e.g., Halo Light) 5.88 3.03 37.3 72.7 11.8 12.1 27.5 9.1 17.7 3.0 Smartwatches/bracelets 0.0 0.0 3.7 5.9 7.4 17.7 37.0 41.2 51.9 35.9

52 3.2.7 Survey Results: WZIT Effectiveness One criterion considered when deciding whether to adopt and implement a technology of any type is the technology’s effectiveness. In order to be of value, technologies must perform as they are intended, meet the needs of the technology owner/user, and provide a positive return on investment. Technologies that are deemed to be not effective, however “effectiveness” is defined by the technology owner/user, will have a remote chance of being adopted or remain implemented. For each listed WZIT, the survey questionnaire asked the participants to rate the extent to which the technology is effective at performing specific functions. The rating scale used ranged from 1 to 5 with 1 = not effective and 5 = extremely effective. The functions listed for each technology (and rated by the respondents) were those for which a TRL equal to 6 or more was associated with the technology. Participants were only asked to rate those WZITs for which they indicated previously in the survey that they are familiar with the WZIT. For example, connected and AVs had TRLs of 6 or more for two functions: warn drivers before an intrusion occurs, and protect drivers after an intrusion occurs. Therefore, those participants who indicated that they were familiar with connected and AVs were asked to rate the effectiveness of connected and AVs with respect to warning drivers before an intrusion and protecting drivers after an intrusion. Table 3.10 summarizes the results regarding the perceived effectiveness of each WZIT in performing the corresponding function(s). Those technologies that are perceived to be at least highly effective (mean rating = 4 or greater) for at least one function in which the technology is deemed ready (TRL = 6 or more) are: • Automated equipment with TMA o Barricade to prevent intrusion (mean rating = 4.10) o Protect workers after intrusion (mean rating = 4.13) • Mobile barrier o Barricade to prevent intrusion (mean rating = 4.41) • Automated flagger with intrusion alert o Detect intrusion (mean rating = 4.07) • UAS for signage o Warn drivers before an intrusion occurs (mean rating = 4.4) • Intrusion alert system with equipment-mounted sensor (e.g., AWARE) o Detect intrusion (mean rating = 4.11) • Intrusion detection with computer vision and ranging (e.g., SmartCone) o Detect intrusion (mean rating = 4.22) o Alert workers during intrusion (mean rating = 4.00) For some technologies that received high effectiveness ratings (e.g., UAS for signage and intrusion detection with computer vision and ranging), the number of participants familiar with the technologies was somewhat low. This result may indicate that while the technology is not well known, it is highly regarded by those who are knowledgeable about it.

53 Table 3.10. WZIT effectiveness (1 = not effective and 5 = extremely effective) Technology Function Min. Max. Mean SD Variance N Connected and Autonomous Vehicles Warn drivers before an intrusion occurs 1 5 3.95 1.15 1.31 19 Protect drivers after intrusion 1 5 3.35 1.37 1.88 17 Automated Equipment With TMA Barricade to prevent intrusion 1 5 4.10 1.23 1.5 41 Protect workers after intrusion 1 5 4.13 0.97 0.93 39 Mobile Barrier Barricade to prevent intrusion 2 5 4.41 0.9 0.81 56 Automated Flagger With Intrusion Alert Warn workers before an intrusion occurs 1 5 3.39 1.36 1.85 31 Warn drivers before an intrusion occurs 1 5 3.24 1.38 1.91 29 Barricade to prevent intrusion 1 5 2.73 1.48 2.2 33 Detect intrusion 2 5 4.07 0.8 0.64 28 Alert workers during intrusion 1 5 3.57 1.12 1.24 28 UAS for Signage Warn drivers before an intrusion occurs 3 5 4.4 0.8 0.64 5 Intrusion Alert System With Equipment-Mounted Sensor (e.g., AWARE) Warn workers before an intrusion occurs 1 5 3.65 1.11 1.23 26 Warn drivers before an intrusion occurs 1 5 3.32 1.43 2.06 25 Detect intrusion 2 5 4.11 0.92 0.84 27 Alert workers during intrusion 2 5 3.84 0.97 0.93 25 Alert drivers during intrusion 2 5 3.46 1.04 1.08 24 Intrusion Alert System With Cone/Barrel-Mounted Sensor (e.g., Sonoblaster) Detect intrusion 1 5 3.36 1.26 1.59 28 Alert workers during intrusion 1 5 3.19 1.14 1.31 26 Intrusion Alert System With Networked Cone/Barrel-Mounted Sensor (e.g., Intellicone) Detect intrusion 1 5 3.47 1.24 1.54 17 Alert workers during intrusion 1 5 3.59 1.14 1.3 17 Intrusion Alert System With Pneumatic Tube Sensor (e.g., WAS) Detect intrusion 1 5 2.88 1.02 1.04 17 Alert workers during intrusion 1 5 2.94 1.06 1.11 17

54 Table 3.10. WZIT effectiveness (1 = not effective and 5 = extremely effective) (continued) Technology Function Min. Max. Mean SD Variance N Intrusion detection with Bluetooth Warn workers before an intrusion occurs 1 3 2.4 0.8 0.64 5 Detect intrusion 3 5 3.57 0.73 0.53 7 Alert workers during intrusion 3 4 3.6 0.49 0.24 5 Intrusion detection with computer vision and ranging (e.g., SmartCone) Detect intrusion 3 5 4.22 0.79 0.62 9 Alert workers during intrusion 3 5 4.00 0.82 0.67 9 Queue warning system with networked cone/barrel sensor (e.g., iCone System) Warn drivers before an intrusion occurs 1 5 3.00 1.41 2 14 Dynamic/changeable message sign and speed enforcement Warn drivers before an intrusion occurs 1 5 3.35 1.22 1.5 52 Wearable lighting (e.g., Halo Light) Warn drivers before an intrusion occurs 1 5 3.02 1.32 1.74 50 Alert drivers during intrusion 1 5 2.79 1.25 1.57 47 Smart watches/bracelets Alert workers during intrusion 2 5 3.90 1.04 1.09 10 Notes: N = Number of respondents who are familiar with the technology; SD = Standard deviation 3.2.8 Survey Results: WZIT Adoption Factors The survey questionnaire asked the participants to give their opinion about different factors associated with the decision to adopt WZITs. Specifically, the participants were asked to rate each factor, based on their personal experience, in terms of the importance of the factor to adopting WZITs. The rating scale ranged from not important to extremely important. The adoption factors included were those identified in the literature review (see Section 3.1). A summary of the ratings for each adoption factor is shown in Table 3.11. The table shows the adoption factors from most important to least important within each factor category based on the mean rating provided by the participants, where 1 = Not important and 5 = Extremely important. In the table, N = Number of respondents who are familiar with the technology; SD = Standard deviation. With respect to technology-related factors, the factor considered most important is the level of resistance to environmental impact (mean rating = 4.66). WZITs must withstand the environmental impacts experienced in work zones. Other technology-related factors received high ratings are: having the required features/technical attributes (mean rating = 4.36), level of technical support required (4.28), ability of warning to alert driver/motorist (4.25), ability of warning to alert

55 worker (4.25), and coverage distance (4.18). In terms of organization-related factors, the top three factors in terms of importance to adoption were: level of technical support available (mean rating = 3.97), level of training required (3.88), and potential level of resistance from employees (3.87). Similarly, the top three factors associated with external impacts are: industry-level change requires technology adoption (mean rating = 3.63), government policy and regulation (3.47), and direct competitors adopt similar technology (3.18). A review of the ratings by category reveals that the respondents perceive technology-related adoption factors as most important to the decision to adopt a technology. The mean importance rating for all 11 technology-related factors was 4.15 (range from 3.69 to 4.66). For the 10 organization-related factors, the mean importance rating was 3.59 (range from 2.81 to 3.97). For the four external-related factors, the mean importance rating was 3.28 (range from 2.82 to 3.63).

56 Table 3.11. Importance of factors to WZIT adoption (1 = not important and 5 = extremely important) Adoption Factor Min. Max. Mean SD Variance N Technology-related Factors Level of resistance to environmental impact (rain, heat, etc.) 3 5 4.66 0.57 0.33 76 Having the required features/technical attributes (warning alert sources, power source, etc.) 2 5 4.36 0.82 0.67 77 Level of technical support required 2 5 4.28 0.9 0.81 76 Ability of warning to alert driver/motorist (motorist comprehension of warning signal) 3 5 4.25 0.73 0.54 75 Ability of warning to alert worker (worker comprehension of warning signal) 2 5 4.25 0.84 0.71 77 Coverage distance 3 5 4.18 0.72 0.51 77 Frequency of false alarms 1 5 4.09 0.82 0.68 77 Number of warning alert sources (visual, audio, haptic, etc.) 2 5 4.09 0.8 0.64 76 Durability (reusable, battery life, etc.) 3 5 4.00 0.72 0.52 77 Level of complexity (ease of deployment, movement, retrieval, maintenance, storage, etc.) 2 5 3.84 0.9 0.82 76 Ability to limit worker exposure to traffic 1 5 3.69 1.14 1.31 77 Organization-related Factors Level of technical support available 2 5 3.97 0.83 0.68 73 Level of training required 2 5 3.88 0.8 0.64 75 Potential level of resistance from employees (responsibility, privacy, etc.) 2 5 3.87 0.91 0.82 76 How quick users will be able to influence colleagues 1 5 3.79 1.03 1.06 76 Potential cost savings from using the technology 2 5 3.75 0.9 0.82 73 Level of compatibility with current processes (impact on traffic control setup, etc.) 1 5 3.68 0.97 0.95 74 Competitive advantage derived from using the technology 1 5 3.49 0.98 0.95 74 Top management degree of involvement (championing the adoption or not) 1 5 3.35 0.87 0.75 72 Cost of labor, equipment, maintenance 1 5 3.32 1.2 1.44 74 Organization culture (receptive to change or not) 1 5 2.81 1.31 1.72 69 Externally-related Factors Industry-level change requires technology adoption 1 5 3.63 0.99 0.98 75 Government policy and regulation 1 5 3.47 0.95 0.91 73 Direct competitors adopt similar technology 1 5 3.18 1.21 1.47 71 Client demand 1 5 2.82 1.33 1.76 72

57 3.2.9 Survey Results: WZIT Implementation Strategies Lastly, the survey questionnaire included a question to obtain the participant’s view about strategies to enhance WZIT adoption. Organizations may pursue a variety of strategies to ensure successful adoption of a promising technology. Those strategies may encompass various activities, adding new/additional resources, providing training, and/or another effort to help ensure wide diffusion and use throughout the organization. The literature review (see Section 3.1) provides a list of strategies for effective adoption and implementation that have been identified in prior research. For each strategy in the table, the survey participants were asked to rate, based on their personal experience, the importance of the strategy to driving successful implementation of WZITs. The rating scale ranged from 1 = not important to 5 = extremely important. Table 3.12 provides a summary of the responses to the question. In Table 3.12, N = Number of respondents who are familiar with the technology; SD = Standard deviation. The results reveal that operations-based strategies are viewed as the most essential to successful WZIT implementation out of all of the strategy categories. Of the five operations-based strategies listed, the mean rating in terms of importance was 3.70 (range from 3.36 to 4.09). The category Standards and Regulatory-based strategies was viewed as the next most important category (mean rating = 3.51, range from 2.97 – 3.87), followed by awareness-based strategies (mean rating = 3.44, range from 3.15 to 3.66) and finally financial incentive-based strategies (mean rating = 3.08, range from 2.67 to 3.4). Out of all of the strategies listed, the strategy that received the highest rating for importance was “support from executive management within the firm/organization” (mean rating = 4.09).

58 Table 3.12. Importance of strategies to successful adoption and implementation of WZIT (where 1 = not important and 5 = extremely important) Strategy Category Min. Max. Mean SD Variance N Standards and Regulatory-based strategies Mandatory work zone safety technology policies and regulations 2 5 3.87 0.84 0.7 75 Better enforcement of work zone traffic control policies after the WZITs have been implemented 2 5 3.80 0.89 0.78 74 More WZIT adoption advocacy by the Federal Highway Authority 1 5 3.51 1.00 1.01 74 Availability of competent and proactive WZIT promotion teams and local authorities 1 5 3.42 0.98 0.96 73 Development of WZIT certification program 1 5 2.97 1.23 1.51 73 Awareness-based Strategies WZIT-related educational and training programs for vendors, contractors, DOTs, and policy makers 1 5 3.66 0.96 0.92 77 Availability to collect information on cost and benefits of WZIT 1 5 3.64 1.18 1.4 75 Publicity through media (e.g., print media, radio, television, and internet) 1 5 3.32 1.14 1.3 76 Creation of public highway safety awareness through workshops, seminars, and conferences 1 5 3.15 0.98 0.95 75 Financial Incentive-based Strategies Strengthened WZIT research and development through additional research investment 1 5 3.40 0.96 0.92 73 Financial incentives for WZIT adoption 1 5 3.35 1.18 1.38 71 Acknowledging and rewarding WZIT adopters publicly 1 5 2.90 1.31 1.7 73 Low-cost loans and subsidies from government and financial institutions to develop and pilot-test emerging WZIT 1 5 2.67 1.20 1.44 69 Operations-based Strategies Support from executive management within the firm/organization 2 5 4.09 0.89 0.79 74 Employee involvement in implementation decision-making 1 5 3.92 0.91 0.83 74 Inclusion of specific technology requirements in contracts 1 5 3.70 0.94 0.88 74 WZIT assessment programs 1 5 3.45 0.96 0.92 74 Availability of institutional framework for effective WZIT implementation 2 5 3.36 0.86 0.75 74

59 3.3 Case Studies As described in Section 2.3 above, the researchers conducted in-depth case studies of the application of promising WZITs in construction work zones. The goal of the case studies was to provide a closer and more holistic study of WZIT implementation within its real-life context. The case studies provide an opportunity to gather practical insights and anecdotal guidance from those DOT and contractor employees experienced in the implementation of the technologies. 3.3.1 Case Study Interviewee and Technology Identification The case study process began by initially documenting those survey respondents who indicated that they are available for a follow-up interview related to the WZITs implemented by their organizations. The last question (Question #35) in the online survey asked the respondent: “Are you available for a follow-up discussion to share your experience using WZITs? If you are interested in sharing your experience, kindly select “Yes” and then provide your name and contact information in the additional survey question. Your name and contact information will not be associated with your responses to this survey, and will be kept confidential.” Those respondents who answered “Yes” were directed to the linked-out survey to collect their name and contact information so that the researchers could contact them. A total of 21 respondents indicated that they are available to be contacted for follow-up discussion. Table 3.13 shows a summary of the responses regarding availability for follow-up discussion regarding WZIT. Sufficiently complete responses are those in which the respondent provided enough contact information (email and/or phone number) such that the researchers could contact them. Table 3.13. Survey responses indicating availability for follow-up discussion Organization All Responses Sufficiently Complete Responses Total # of States* Total # of States* State Departments of Transportation (DOTs) 17 15 17 15 Industry (contractors, subcontractors, WZIT manufacturers/vendors, other) 4 4 4 4 Total 21^ 21 * Some states may be common between DOT and industry responses. ^ An additional four responses were received; however, the responses did not contain any information. The researchers aimed to conduct four to six case studies of individual WZITs or types of WZITs. The process of selecting which technologies to investigate in the cases studies began with the results of the TRA. Those technologies with TRLs equal to 8 and 9 were initially identified. The researchers also performed a preliminary review of the online survey responses to identify the technologies that are commonly implemented in current practice. Lastly, the researchers looked at the survey responses from those states in which a respondent indicated availability for a follow-up interview. This analysis resulted in an initial list of 13 technologies to target for the case studies. The shortlisted technologies were:

60 • Connected and autonomous vehicles • Automated equipment with TMA • Mobile barrier • Automated flagger with intrusion alert • UAS for signage • Intrusion alert system: o equipment-mounted sensor (e.g., AWARE) o cone/barrel-mounted sensor (e.g., SonoBlaster) o networked cone/barrel-mounted sensor (e.g., Intellicone) o pneumatic tube sensor (e.g., WAS) o computer vision and ranging (e.g., SmartCone) • Queue warning system with networked cone/barrel sensor (e.g., iCone System) • Dynamic/changeable message sign and speed enforcement • Wearable lighting (e.g., Halo Light) The researchers also reviewed the HSM (AASHTO 2010) to determine whether those technologies on the shortlist are addressed in the HSM and, if so, record the guidance provided by the HSM for implementation of the technologies. The HSM content that relates to work zones is summarized below: • Section 16.4 – Crash Effects of Work Zone Design Elements: o This section provides crash modification factors (CMFs) associated with work zone design elements. The available CMFs included in the HSM are presently limited, and CMFs are only provided for modifying the work zone duration and length. • Section 16A.3 – Work Zone Design Elements: o This section describes three different elements of work zones that may be included in the design of a work zone. The elements described are:  16A.3.1: Operate Work Zones in the Daytime or Nighttime  16A.3.2: Use Roadway Closure with Two-Lane, Two-Way Operation or Single Lane Closure  16A.3.3: Use Indiana Lane Merge System (ILMS) • Section 16A.4.1 – Work Zone Traffic Control and Operational Elements: o This section provides general information and descriptions of several types of work zone traffic control that are included in the MUTCD, as well as currently identified trends related to crashes and driver behavior in work zones. o Section 16A.4.1 – General Information:  Signs and signals  Delineation  Rumble strips o Section 16A.4.2 – Trends in Crashes or User Behavior for Treatments with No CMFs:  Install changeable speed warning signs  Install temporary speed limit signs and speed zones  Use innovative flagging procedures  Install changeable message signs  Install radar drones

61  Police enforcement of speeds • Section 16A.6 – Treatments with Unknown Crash Effects: o This section provides a list of treatments for which the impact of the treatment on crash rate is not known. o Section 16A.6.2 – Work Zone Design Elements:  Lane closure design  Lane closure/merge design o Section 16A.6.3 – Work Zone Traffic Control and Operational Elements:  Signs and signals  Delineation  Rumble strips  Speed limits and speed zones As can be seen from the description of the HSM content above, the HSM contains limited information about technologies used in work zones, especially those technologies that are designed for preventing and mitigating work zone intrusions. In addition, the information provided does not specifically address use of the technologies for preventing/mitigating work zone intrusions. Those technologies mentioned in the HSM that are recognized as potential WZITs are: temporary rumble strips, changeable message signs, drones (UAV), and drones equipped with radar for speed monitoring and enforcement. As shown in Table 3.13 above, a total of 21 respondents indicated their availability for follow-up discussion to provide detailed technology information sufficient to develop the technology case studies. Based on the shortlisted technologies, and the list of respondents who were available for follow-up discussion, the researchers initially identified six states to contact to collect technology documentation for the case studies. The selected states, along with the technology(ies) targeted in each state, are as shown in Table 3.14.

62 Table 3.14. States targeted for follow-up discussions State Technology(ies) Alabama • Dynamic/changeable message sign and speed enforcement Colorado • Connected and autonomous vehicles • Intrusion detection with computer vision and ranging (e.g., SmartCone) • Queue warning system with networked cone/barrel sensor (e.g., iCone System) California • Connected and autonomous vehicles • Automated flagger with intrusion alert Texas • Mobile barrier • Automated flagger with intrusion alert • UAS for signage Minnesota • Intrusion alert system: o pneumatic tube sensor (e.g., WAS) Virginia • Automated flagger with intrusion alert • Intrusion alert system: o cone/barrel-mounted sensor (e.g., Sonoblaster) • Dynamic/changeable message sign and speed enforcement • Wearable lighting (e.g., Halo Light) Oregon • Mobile barrier • Intrusion alert system: o cone/barrel-mounted sensor (e.g., Sonoblaster) o networked cone/barrel-mounted sensor (e.g., Intellicone) o pneumatic tube sensor (e.g., WAS) Each case study developed focuses on a different technology. Where information about a technology was collected from multiple states, the information was combined into one case study. The information collected for each case study included: technology characteristics, application and usage details, implementation plan, and user evaluation of WZIT effectiveness. Other details of interest, such as the typical environmental conditions of usage, frequency of usage, applicable project characteristics, and other pertinent information specific to the case study under investigation, were also collected. These data were collected through a multi-modal data gathering effort consisting of the following sources and methods: semi-structured interviews, review of manufacturer documentation and archival literature, and/or observation of WZIT implementation. 3.3.2 Case Study Interviews The interviews targeted one or more technologies being implemented by the organization. An initial set of interview questions was developed to gain further information about each technology and maintain consistency between interviews. The interview questions related to each technology were as follows: 1. Why did you choose to implement the technology? 2. What did you use before using the technology?

63 3. How is use of the technology better/worse/different? 4. On what types of projects has the technology been deployed? a. Highway or other roadway projects? b. Nighttime or daytime work? c. Any specific type of work activity? 5. Do you have any notable experiences to share about the technology? 6. Please rate your satisfaction with the technology from 1 to 3, where 1 = low and 3 = high? 7. Using a scale of 1 to 3 where 1 = low and 3 = high, please rate the usability of the technology for each of the following: a. Deployment b. Retrieval c. Movability with the site 8. Using a scale of 1 to 3 where 1 = low and 3 = high, please rate the maintenance requirements of the technology for each of the following: a. Time b. Effort c. Cost 9. Would you recommend the technology to others? 10. Are there any limitations associated with the technology, or improvements in the technology that you would like to see? 11. Are there any lessons learned based on the technology implementation and use? 12. Is there any other project and/or technology information that you could share with us that we can include in our case study (e.g., project and/or technology photos, project documentation, figures/tables, etc.)? 13. Do you have any other general suggestions or recommendations regarding the use of WZITs on roadway construction projects? 14. We will be developing a descriptive case study(ies) based on the information that you have provided. a. Is it OK that we use the information that you provided in our case study? Your name and personal information will not be included. b. If we have additional questions as we develop the case study, may we contact you again? The researchers also asked additional probing questions to gather more details about the technologies where warranted. To facilitate and streamline the interviews, the researchers developed a script for conducting the interviews. The script was followed by the researchers to ensure coverage of all needed items and consistency between interviews. The script used for the interviews is provided in the Appendix. To schedule the interviews and gather information about the targeted technology(ies), the researchers contacted each state. Initial contact was made via email to re-confirm their interest and availability for an interview, and to establish a date and time to conduct the interview via an online meeting program (e.g., Zoom). The interview was then conducted at the time were confirmed.

64 The research team interviewed personnel in seven state DOTs and one contractor to gather critical information on the following six WZITs: 1. AFAD 2. Mobile barrier 3. ATMA 4. Smart work zone (queue warning system, DMS, and speed enforcement) 5. IAS 6. Wearable lights The DOTs in the following states participated in the interviews: Alabama, California, Colorado, Minnesota, Oregon, Texas, and Virginia. In certain cases, the researchers conducted multiple interviews for a state DOT to ensure relevant and reliable information was gathered. For instance, the researchers interviewed two groups in Caltrans and Alabama DOT to gather information on different technologies used by these states. 3.3.3 Results: WZIT Case Studies The technology and implementation data collected as part of this task was developed into formal case studies. The researchers initially developed a preliminary outline of the case studies that was used as a template for documenting the case studies. Each case study focuses on a specific WZIT technology and its implementation in practice. The template provides a common format for each case study for consistency in case study documentation and to organize the material to support case study analysis and reporting. The initial case study template, provided in the Appendix, includes sections for the following content: • Technology deployed • Technology functions used • Applications • Technology adoption • Pre-planning and implementation plan • Observed impacts and effectiveness • Barriers to use and limitations to effectiveness • Recommendations for use • References and additional information When preparing the final case studies, deviations in the template were made if warranted based on the technology information collected. Essential information was extracted from the interviews and supporting documents for the six WZIT case studies. In some instances, the case study focuses on a single technology (e.g., automated flagging assistance device), while other case studies focus on a suite of similar technologies (e.g., IAS). Summary descriptions of the case studies are provided below for brevity. The detailed case studies for each WZIT are provided in the Appendix.

65 3.3.3.1: AFAD AFADs are remotely operated TTC equipment with high visibility signage that help to either control the access points of a work zone or direct and control the traffic flow (https://www.northamericatraffic.com/products/flagging-devices/rcf2-4-automated-flagger- assistance-device). A typical flagger system consists of the use of either one or two AFAD units. An AFAD unit is located at the entrance to a work zone and uses either a slow/stop paddle or a red/yellow lens to inform the drivers when it is safe to pass through the work zone, along with a gate arm to limit passage into the work zone. Caltrans has implemented this technology, and those Caltrans employees interviewed as part of the study highly recommend that other DOTs implement AFADs as well. The information provided below comes primarily from the interview of Caltrans employees, along with information about AFADs that is publicly available online. Caltrans has utilized this technology on a 2-lane highway to control traffic flow around culvert, drainage, and slug removal work operations. Caltrans targeted using the Automated Flagger AF- 100 (http://noflaggers.com/Page22/Default.htm), which is a type of AFAD that is compatible with the operation and motorist compliant. The AF-100 is a portable traffic control device that meets the 2009 FHWA MUTCD Section 6E.04 specification. The device used has a gated arm along with red/yellow lens and is used in single- or two-trailer operations. The combination of the red/yellow signal lights and gate arm is visible to oncoming traffic to instruct the motorists when it is safe to proceed into the work zone. AFAD systems should be operated by qualified flaggers who have been trained in the use and operation of the type of AFAD system to be used for the work zone. AFADs are typically selected for use in the work zones that have higher rates of close calls, near misses, minor crashes, and low- risk non-compliance incidents. Currently, in California, a single AFAD is given to a crew unit and Caltrans is looking to acquire and distribute a greater number of AFADs for work crews and expand their use. Caltrans employees interviewed recommended the use of red/yellow lens style AFADs in cities where the population includes a high proportion of drivers who do not understand the slow-stop paddle sign style AFAD, such as tourists from other countries. The AFAD was used and implemented by Caltrans based on the guidelines suggested by the manufacturer. Two separate guidelines were suggested, one for a single flagger/operator and the other for a two flagger/operator operation. Further guidelines are suggested when the AFAD system is inoperable or not in use. The particular AFAD system selected was operated remotely and met the requirements of the MUTCD. It also needed a special hook device that would attach the AFAD to the truck. Caltrans employees interviewed expressed a positive reaction to the use of a red/yellow lens with gated arm style AFAD. Those interviewed felt that it is better and less difficult to use than the traditional flagger operations and/or slow/stop paddle type AFAD. Confusion among motorists regarding whether to either slow down or stop, which is a motorist compliant, was reduced. The reason that the red/yellow lens was used was to ensure that it is visible to motorists. As a result, Caltrans decided to develop specifications for a specific style of AFAD to be used in work zones

66 throughout the state, where applicable. The flagger operator had to manually press a button to sound an alarm during a vehicle intrusion incident. In a few incidents, the minor crashes and close calls were averted and, in a few cases the AFADs got hit by the intruding vehicle and the gated arm and/or the system hub was destroyed. In such cases where AFADs have been destroyed, the units were replaced by human flaggers. Those Caltrans employees who used the AFAD indicated that the unit was heavy in weight and difficult to move around the work zone by hand. A pickup truck was needed to make multiple trips to move around the AFAD from one direction of the road to the other. This issue was of main concern along with the fact that the alarm system requiring manual activation did not allow the workers to respond to the incident in time. Caltrans suggests there should be more updates expected from the manufacturers such as an auto alarm system powered by sensors or lidar. The Caltrans employees interviewed recommended using AFADs for projects where there is a high average daily traffic (ADT) level or high number of road users. Those interviewed recommended placing the AFAD based on the available project safety data and previous project experience, and using the AFAD based on the type of work zone activity [for example: culvert work, cleaning drainage, slug removal, and crack sealing]. Several suggestions for enhancements to the AFAD were received through the interviews, including: addition of a smart camera to detect intrusions, addition of an automated device such as a LiDAR can keep track of the physical conditions of the AFAD, and use of sensors to deploy invisible beams or rays that would signal the alarm system when a vehicle breaks the line of sight. Interviewees also felt that AFADs are more compatible with long-term work zone operations, such as paving a 5-mile length of roadway. To tackle the issue of weight, Caltrans is following research conducted by the Minnesota DOT where Minnesota DOT is utilizing wheels to help move the AFAD around the work zone. Another recommendation provided is the use of an intrusion alert device, such as SonoBlaster, Intellicone, and WAS, placed near and strategically aligned with the AFAD. Using a combination of devices in this way is expected to improve safety for the motorist and help decrease the number of intrusions. However, the drawback to this form of setup is that it is costly. Based on the assessment of the technology, Caltrans employees interviewed believe that, compared to the cost, the worth of the technology lies in its functionality and the reduced worker exposure to the passing vehicles. 3.3.3.2: Mobile Barrier A mobile barrier is a mobile traffic barrier system used to provide positive protection between passing vehicles and the work area. Mobile barriers serve to prevent intrusions into a work area and therefore protect the workers from vehicle intrusions. Mobile barriers are easy to set up and remove in and around short-term work zones during peak traffic hours. Mobile barriers are useful during both the daytime and nighttime. Some mobile barriers have lighting systems to illuminate the work area and a dynamic message sign attached. A crash attenuator may or may not be attached to provide a cushion against the impact of vehicles from the rear end of the truck. The MBT-1®, developed and manufactured by Mobile Barriers, provides 42 to 102 feet (12 to 31 meters) of highly mobile positive barrier protection with minimal deflection (https://www.mobilebarriers.com/media/docs/Mobile%20Barriers%20Products%20and%20Opti ons%20-%200618.pdf). It can be used with any standard semi-tractor with 60 inches (1.5 meters)

67 of swing clearance. The barrier can be transported fully set up at freeway speed from site-to-site. It is commonly used with a cabover or day cab tractor. Configured at 62 feet (18.9 m) using one wall section, the MBT-1® drives similar to a semi-trailer. The MBT-1® can be easily reconfigured to the right or left lane closure configuration depending on the end to which the tractor is attached. The MBT-1® features lockable and on-deck storage and can be ordered with integrated electrical power, lights, signage, TMA, crane, and other features. An MBT-1® is available for sale, lease, or rent (https://www.mobilebarriers.com/resources-drivability.html). The Mobile Barriers MBT-1® was developed to help improve work zone conditions for workers and the public (Mobile Barriers, n.d.). It is, in essence, a portable, self-contained, work zone that simply drives in place much like a semi-truck, and then can be transported off the site as quickly when the work is done. It is designed to provide a safer, better work environment for workers, and minimize disruption, improve traffic flows, and allow for reopening lanes more quickly for the public. Mobile barrier use is seen especially for highway construction, maintenance operations, utility work zones, and very short-term work zones where fast-flowing traffic is present and the workers have very little to no protection available. The work zone configurations may be setup for long-term or short-term durations. The MBT-1® provides a wall of steel adjacent to the work area that prevents vehicle intrusion into the work area (Gambatese and Tymvios, 2013). The Oregon Department of Transportation (ODOT) uses a mobile barrier for long-term and short- term work zones involving highway construction, maintenance operations, and utility work. ODOT employees interviewed indicated that ODOT has two mobile barriers available that are primarily used by for bridge maintenance operations. The bridge crew configured a way to use the mobile barriers where one is placed at an optimal location nearby work zone and the other one is used at a high-risk location nearby, such as when there is a horizontal or vertical curve located right before the bridge. ODOT uses the mobile barriers for the following work zone operations involving surface, drainage, traffic services, and structural operations: minor/major surface repair, deep base repair, concrete repair, crack sealing, minor culvert and inlet cleaning, minor culvert and inlet repair, pavement marking, major/minor sign installation maintenance, traffic signal maintenance, illumination maintenance, guardrail/barrier maintenance (repair or clean), flasher/beacon maintenance, guardrail/barrier maintenance repair and clean, attenuator maintenance, bridge maintenance and repair, and lastly, structural painting operations. The ODOT interviewees mentioned several planning issues that had to be incorporated when deciding when and where the mobile barrier was to be set up: A truck with a TMA (“crash truck”) is used as a supplement to the mobile barrier during static lane closures as a means of positive protection to mitigate head-on collisions with the vehicle. When a static lane closure is implemented, a mobile barrier and cones are used depending upon the project work specifications. A truck with a TMA along with mobile barrier is placed partially on the road and in shoulder lane. The use of an arrow board indicates to the drivers the need for a lane change. For ODOT, those ODOT employees interviewed indicated that cost and the time constraints were important critical decisive factors of concern when the budget available to be spent on work zone technologies was considered. Cost and time are also considered when determining whether to deploy a mobile barrier on a project.

68 Crash testing of the MBT-1® conducted by the manufacturer showed the mobile barrier moving only inches away from its position when struck by an oncoming vehicle during the test. The ODOT employees felt that this result was a good sign of faith in the technology. The mobile barrier system was evaluated using the following performance metrics: time required to set up the barrier at the start of the operation and demobilization of the system at the end of the operation: limitations/enhancements to work operations as a result of using the mobile barrier; worker safety and safety perception; worker productivity while the mobile barrier was present during the work activity; motorist safety and safety perception by the road users; mobile barrier system performance based on project/site attributes; and the ease of transportation of the mobile barrier to/from the work zone. The Minnesota Department of Transportation (MnDOT) utilizes a mobile barrier for some of its work and those MnDOT employees interviewed felt that the mobile barrier is functional for use during approach panel type work zone activity. A sense of complete protection is felt by the MnDOT personnel in some of work zone projects that utilize the mobile barrier. MnDOT will buy add-ons if the manufacturer adds additional safety components. In terms of productivity, ODOT employees interviewed did not find the mobile barrier productive in locations different that small spaces. Use of the mobile barrier is seen as functional for a small area involving long-term duration work, and also is compatible with the work performed by the bridge maintenance crew. ODOT highway construction personnel did not feel comfortable reconfiguring the mobile barrier (i.e., changing the barrier configuration from one side to the other) during the same working day. Physical limitations associated with the barrier were faced by the bridge maintenance crew. Movement of mobile barrier in the presence of live traffic was difficult. A need for pre-planning is required in order to close a lane for the work zone duration. The truck had to reconfigured (i.e., moved to the other end of the barrier) when changing the side in which the work was going to take place, which caused the trailer to be relocated. In some cases, the overhanging beam in the mobile barrier requires the workers to close two lanes instead of one lane. Due to the use of a static (non- moving) lane closure, the work carried out is less. In addition to the traditional equipment and tools that the crews are equipped with, and that are required by the standards and guidelines, ODOT currently owns and operates a Quickchange Movable Barrier System that uses concrete reactive tension system (CRTS) blocks. Use of the additional equipment shows promising results as per the projects in the Portland, Oregon area (Gambatese and Tymvios, 2013). According to the MnDOT personnel interviewed, the need for a truck to move around the mobile barrier is seen a huge downside as the manufacturers sell the standalone mobile barrier. Space to switch out the axle connections is sometimes difficult to find. In addition, the need for the crew to pre-plan the use of the mobile barrier on one side of the work zone and finish that part of work before switching the mobile barrier to the other side for work on the other side of the barrier is seen as cumbersome. Furthermore, the purchase cost (approximately $330,000) was also viewed as an issue as MnDOT faced a time constraint to include the mobile barrier in the budget and use it efficiently. An issue arose as well as a result of the very large size of the mobile barrier. Due to

69 its large size, the barrier utilizes the full work zone lane and protrudes into the adjacent lane during the work zone operations. ODOT has experienced cases in which passing vehicles side swiped the mobile barrier when it is present on a roadway with narrow lanes. In another case, the mobile barrier also was struck as the through lane narrowed due to the presence of a guard rail. On one project, the contractor was unable to incorporate the cost of renting a mobile barrier in its bid, so ODOT bought one and offered it to the contractor for use. MnDOT employees interviewed recommend use of a mobile barrier after further safety equipment additions or modifications and suggested changes are made. ODOT professionals interviewed recommend use of a mobile barrier for a small area that involves long duration work. 3.3.3.3: ATMA TMA, also referred to as an impact attenuator and crash attenuator, is a device mounted to a truck that is designed to reduce the damage resulting from an errant motor vehicle collision into the truck (https://www.kratosdefense.com/systems-and-platforms/unmanned-systems/ground/autonomous- truck-mounted-attenuator). TMAs are commonly used for protection of workers during many different types of roadway construction operations. An ATMA is an autonomous truck with a TMA. As an autonomous vehicle, the ATMA uses advance object detection and communications to allow it to follow a lead vehicle autonomously, i.e., without a driver. This capability removes a worker from being exposed to potential crashes while still providing protection to the lead vehicle. Kratos Defense, along with Royal Truck and Equipment pioneered the field of “Highly Automated Work Zone” vehicles, by deploying the world’s first ATMA, also known as an autonomous impact protection vehicle (AIPV). The unmanned ATMA/AIPV concept was leveraged from systems Kratos Defense developed for the U.S. Military to reduce warfighter support of dangerous missions by converting human-driven convoy vehicles into unmanned systems. Some state DOTs, like Caltrans and Colorado Department of Transportation (CDOT), are utilizing ATMAs in their construction and maintenance operations. In an interview with Caltrans employees, the interviewees stressed the recent developments in the technology, and expressed that the ATMA is still being developed. Manufacturers are making adjustments to the systems used by Caltrans, with upgrades according to issues raised by Caltrans. The ATMA technology used by Caltrans and other similar systems have a variety of capabilities. Kratos Defense, for example, has automated fleet vehicles using a multi-platform applique kit (M- PAK) to enable the self-drive feature in multi-vehicle leader/follower operations with a human- driven highway maintenance vehicle (i.e., line striping or road sweeper vehicle). In the leader/follower configuration system, the maintenance (lead) vehicle transmits navigation data via encrypted vehicle-to-vehicle (V2V) communications to the ATMA (follower). The trailing vehicle uses the data to follow behind the lead vehicle. The following vehicle is completely unmanned and follows along the route taken by the lead vehicle. The ATMA system features component redundancy, an active safety system, high accuracy GPS/GPS-Denied navigation, encrypted V2V communications, front and side view obstacle detection, and a robust user interface which provides system feedback, situational awareness, multi-camera view, and operator control.

70 Caltrans employees interviewed suggest the best application for ATMA use is paint striping. The CDOT also uses an ATMA to prevent rear-end crashes during paint striping operations. Caltrans professionals interviewed indicated that the ATMA used was a trial version and recommended waiting until the ATMA technology is completely developed or has fewer operational issues. The Caltrans members expressed major concern about the TMA fleet vehicle being able to take the impact of vehicles on high-speed routes which have speed limits ranging from 65 mph to 75 mph such as a state route or interstate highway. On the other hand, the ATMA vendor made sure to provide detailed extensive training where the system components were tested and validated. A training manual was provided, but the users still raised safety issues to the manufacturer. The training consisted of in-cab training related to the ATMA operational features such as the “E-Stop” and “A-Stop” features. E-Stop, short for Emergency Stop, is a button that the lead vehicle driver can press when an obstacle is present that could not be crossed over, or during the event of an emergency. When the E-Stop button is pressed, a signal is sent from the lead vehicle to the following vehicle to direct the following vehicle to stop. “A-Stop” is short for Automated Stop, and is used when the lead vehicle system detects obstacles present in the front of the vehicle, or observes a large obstacle that the lead vehicle cannot pass through. In this case, the lead vehicle also communicates with the follower vehicle to stop. The drawback of the E-stop is that the system has to re-boot manually while the ATMA vehicles remain stationary until the operations system is functional. The DOT employees interviewed shared their experiences using the ATMA in terms of its impacts and effectiveness. Provided below are both advantages and drawbacks of the systems used. Interviewees believe that the ATMA provides better lateral control accuracy and it exceeds more than what it is expected to do. However, in some cases, the radar present in the front section of the ATMA was somewhat nullified or did not account for the curves on the roads while following the leader vehicle. When an errant vehicle impacted the ATMA, the impact of the collision with the TMA was much more dramatic on roads having speed limits up to 75 mph as compared to the roads having a speed limit of 35 mph or less. Collisions that occur in high-speed situations are expected to be more serious in nature when it comes to injuries than low-speed collisions. The ATMA control system used for positioning, navigation, and communication requires re- booting in various circumstances. In some cases, the wait time for re-booting the system can be lengthy. However, with recent system upgrades, the wait time is significantly reduced from five minutes to 15 to 20 seconds. Similarly, an issue with alignment during the system re-boot is being fixed with upcoming system updates. The cost-benefit trade off can be seen as a concern, but the features and application can be justified for the cost. The Caltrans professionals interviewed addressed a concern that if there is an incident in which the lead vehicle is involved in a crash, the workers and drivers will be out of work for a period of time. Worker union issues are also a factor that should be considered. A driver has to be employed specifically to take care of the ATMA while it is operational in the work zone boundaries and to drive the ATMA to and from the work zone operation site. California State laws do not allow the use of “AVs,” however this regulation is not stressed right now as the ATMA is still in its research

71 and development stage. The base purchase cost of the ATMA was projected to be approximately around $330,000, which increased to $410,000 with the upgrades. The Caltrans employees interviewed mentioned that during one of the ATMA tests on a closed section of highway, the ATMA lost connection with the leader vehicle when passing underneath a bridge. The interrupted connection was due to a loss of connection with the GPS, and the inertial navigation was activated. However, the ATMA swerved pretty dramatically as it got reconnected with the lead vehicle via GPS. In such instances, the ATMA system uses gyros and/or wheel encoders in combination and the system integrates a forward projection as a solution. The forward projection is precise for a short distance and time, but will tend to drift off in the long run. The ATMA vendor company has taken into account the curvature of the road not being detected by the ATMA. The identified issue has been resolved and the ATMA has been updated with a new radar- LiDAR hardware component in the front of the vehicle which takes into account obstacle detection with a field of view (FOV) of 180 degrees. In 2020, there was a major update to the ATMA software systems and the hardware was updated with new sensors. New tests will be conducted to verify that the identified issues have be fixed. For example, with respect to the issue associated with loss of connection when passing under a bridge, the issue was deemed resolved when the ATMA system passed in 12 out of 15 tests. These tests indicated the system is functional. Following the new upgrade, the underpass test can be avoided or an additional five tests could be run as the ATMA system has already been tested successfully during the previous round of tests. Maintenance of the operating system will be reduced as ATMA vendors upgrade their models. Caltrans has successfully tested a third version, and is waiting to receive the fourth version. Conversely, maintenance of the hardware will be an issue and could be worked out as an agreement with the manufacturer or a partner company. The DOT employees interviewed recommended using the ATMA for paint striping where the ATMA vehicle follows the paint truck (lead vehicle) at a projected speed of 10 to 20 mph. However, moving forward with using the ATMA in this way at this time requires resolving operational issues through system updates or installing new hardware components. One of the maintenance divisions of a district said that they are interested in trying out the ATMA. Apart from paint striping, maintenance personnel indicated that the ATMA also shows excellent potential application for pot hole filling operations. 3.3.3.4: Smart Work Zone (queue warning system, DMS, and speed enforcement) A smart work zone, also known as an intelligent work zone, is a site-specific configuration of traffic control technology installed within a roadway work zone to improve construction worker safety, give “real-time” travel information to the road users, and efficiently route vehicles through a work zone. Smart work zones perform numerous manual functions automatically such as eliminating the need for human “flaggers,” thereby removing such workers from exposure to hazardous situations. Smart work zones typically integrate the use of numerous disparate technologies such as sensors for determining vehicle speed and intrusions, as well as output technologies like DMS. Examples of smart work zones that are discussed here include queue warning systems (QWS), DMS, and truck warning systems.

72 • QWS alert vehicle drivers to an impending traffic queue by transmitting electronic messages to PCMSs and providing warnings to vehicle drivers at the upstream location of the construction zone. Another feature is that it alerts vehicle drivers to slow down owing to traffic ahead, or to take a detour to save time before reaching the starting location of the traffic queue. A QWS is recommended for use on long-term interstate roadway projects where stopping sight distances are not adequate for drivers to observe the formation of queues near work zones. Examples of such instances include work zones located in roadway sections containing curved horizontal alignments and changing vertical alignments. These systems are also particularly useful in roadways with high traffic volumes. While a queue warning system is not designed as a work zone intrusion device, its use in connection with other technologies could perhaps provide warning of a potential intrusion. Truck warning systems are very similar to QWS and serve to warn the passenger vehicle drivers that a truck is entering and/or exiting the work zone, and advise them to be vigilant and to perhaps move over to a different open lane. These systems are also most applicable when visibility of the work zone and stopping sight distances are an issue due to road and weather conditions. • DMS enable the provision of customizable and dynamic messages for drivers as they approach work zones. DMSs have been proven to be effective in lowering mean vehicle speed before the work zone through a variety of changes such as attaching a radar unit to the message sign to notify drivers of their speed. As an example, ALDOT used a DMS to change messages and attract the motorist attention to be aware of an upcoming work zone around a bridge. Of note, the bridge was in a region with roads and a lane closure right before roadwork on the bridge. The interviews with Colorado and Alabama DOT personnel provided useful advice for adoption and implementation of the QWS and DMS technologies: • QWS: The Colorado DOT (CDOT) conducted a study to identify the approximate length of the queue to determine the number and positions of traffic sensors and DMS. When slow or stopped traffic is detected by sensors, a warning message is triggered on the DMS board. Sight distance is seen as an important factor where the visibility of the end of queue is necessary for the motorists to brake before the queue is visible. It was noted that initial costs are very high to implement the QWS. • DMS: The Alabama DOT (ALDOT) noted that one of the prime considerations for implementations was the use of consistent abbreviations due to limited space for displaying the full message on the board. Use of the DMS also required planning by staff designers or consultants in terms of location to be placed at, especially given that use of the DMS required a high upfront investment. Use and assessment of the systems has provided preliminary insights into their impact on vehicle speeds and intrusions, and on their effectiveness at preventing crashes. The interviews with Colorado and Alabama DOT personnel provided information in this regard: • QWS: CDOT experienced high levels of accuracy when setup factors such as the location and alignment of devices were optimal. Use of the QWS was noted to be a labor-intensive process and training for the contractors was important. The performance of the QWS was

73 validated through field observations, available data, and travel time data. The QWS fared well in all aspects and functioned as expected. The QWS was seen to be functional for 90% of observed time. Due to movement of a traffic alignment ramp, the portable QWS had to be moved a lot and realignment of the devices was repeatedly suggested to the contractors. • DMS: ALDOT provided favorable reviews of the use of the DMS for their smart work zone application and noted that the public was aware of the upcoming road conditions and that the technology helped to avoid high-speed crashes and ease the traffic congestion. Constant attention towards the device conditions and alignment is suggested by those CDOT employees interviewed. Furthermore, system updates were required every two months. During the system update, the QWS was inactive and there were higher chances of the rear/end crashes and impacts. CDOT engineers interviewed observed that the drivers did not slow down during the curved section of the road even though they saw a DMS. A QWS cannot be implemented in work zones around bridges where the space is limited and physical limitations are present. This limitation needs rectification and was suggested by CDOT as the communications systems had some network issues such as signals from the sensor to the DMS not being received. A training program for the workers is suggested to ensure the devices are aligned and functional. The QWS is recommended for use in long-term work zones on interstates where curved sections and different elevations of the roadway are found. The QWS can be employed in traffic environments where long queues may occur, and to provide accurate and actionable traveler information to the public, reduce traffic delay through work zones, improve safety of traffic merges in work zones, improve incident management and emergency response, improve compliance with reduced work zone speed limits, decrease queue lengths associated with work zone congestion, lessen secondary crashes in queues, improve driver compliance where there are vehicle backend/rear collisions, and diminish the disruption of traffic flow to the road users. 3.3.3.5: Intrusion Alert Systems (IASs) IASs are generally single or multiple WZITs that are integrated into a system to provide feedback to the workers through audio, visual, and/or vibrational means about a vehicle intrusion and ensure that a highway worker has sufficient reaction time to protect themselves from the intruding vehicle. Table 3.15 lists various IASs along with their primary technology components and output warnings produced. Table 3.15. Summary of IAS Intrusion Alert System Type of Sensor Technology Used Audio Alert Visual Alert Vibrational Alert Traffic Guard WAS Microwave and pneumatic tube-based Present Present Present SonoBlaster Kinematic energy Present Not present Not present Intellicone Radio wave signal Present Present Not present AWARE* Radar Present Present Present *The technology is still in its final development stages.

74 The primary function of an intrusion alert system is to alert workers of an intruding vehicle that occurs before, during, or after the intrusion across the work zone boundary. The technologies have audio alert, visual alert and vibrational alert features, and are used as a standalone device or in combination with other devices to warn the workers of an intrusion. Each IAS is described in more detail below: • The WAS is an auditory and visual alarm which wirelessly triggers when a vehicle passes over the pneumatic tube and activates strobe lights, an audible alarm, and a vibrational alert message to the workers. • The SonoBlaster is a system that emits an audible alert sound which is activated when the cone/barrel on which the unit is placed is impacted/struck by the vehicle. The SonoBlaster consists of a CO2 cartridge and alarm unit. When the cone/barrel is struck and the SonoBlaster is tipped over, the SonoBlaster emits an air horn sound. • The Intellicone is a portable site alarm (PSA) which is powered by motion sensitive and communications-based technologies. The device utilizes General Packet Radio Service/Global System for Mobile (GPRS/GSM) communications and GPS sensors. • The AWARE system is a radar-based system that can identify a potential work zone intrusion from several vehicles while also warning the errant driver and personnel who may be in danger. The system is comprised of a sensor that includes electronically-scanned radar, high-precision differential GPS, accelerometers, gyroscopes, and magnetometers for position and orientation sensing. Installation of the IASs is fairly simple and can be done by a single worker. Each system has its own steps to follow for installation and application. The following are recommended installation steps and device applications: • WAS installation is done manually by the operator by laying out a pneumatic hose transverse to the flow of traffic. The steps to deploy the unit are as follows: First, the trip hose is deployed across the lane and the hose pressure sensor box is powered up by pressing a button. Second, the portable alarm case (PAC) is attached to a piece of infrastructure on the roadway inside the work zone. Third, the workers turn on their PSD and verify the green LED is visible. Lastly, the pneumatic hose is stepped on to trigger the alarm and test the system for operability. • A SonoBlaster is recommended for long-term work zones. The CO2 cartridges should be tested for their functioning, and if not operational, the crew should have extra CO2 cartridges as a standby. • The Intellicone system devices can be mounted on a traffic cone or channelizer using bolts. The electronic system can be mounted on cones and transmits the signals when impacted by a vehicle to produce auditory and visual alarms in the work zone. Gambatese et al. (2017) noted that the WAS technology strengths are “ease of use” and “effectiveness of triggering mechanism.” The cost and effectiveness of the device’s alarm, on the other hand, were cited as disadvantages of implementing the WAS technology. The decibel level of the alarm is an important concern given the noise present in the surrounding work zone. For example, heavy construction equipment typically emits sound levels that range from 80 to 120 dB. The noise level at a distance of 25 feet from a passenger vehicle moving at 65 mph is approximately

75 77 dB. To be effective, the alarm must emit a sound that is louder, and distinctive from, the surrounding noise. The warning sound is 110 decibels at 0 feet and gradually decreases to around 85 dB at 50 feet. Alarm decibel levels do not fluctuate significantly from around 85 decibels at 50 feet to approximately 65 decibels at 300 feet. The radio wave transmission rate between the sensor and PSA is effective up to 100% at 400 feet and 0% at 500 feet. The researchers noticed that the worker’s maximum median reaction time was roughly 1.22 seconds in response to an alarm with a mean duration of 4.7 seconds when the vehicle contacted the sensors at 30 mph. In a study that tested the SonoBlaster, its cumulative response rate for three tests were 92% and 85% at 50 and 100 feet, respectively. The audio level of the SonoBlaster can be adjusted and is seen as a positive point by Caltrans professionals interviewed. In another scenario, the Virginia Department of Transportation (VDOT) practitioners interviewed expressed that the SonoBlaster was rendered useless as it was run over by a vehicle and the CO2 cartridge punctured without sounding the alarm. In other narrow work zone sites, the SonoBlasters were knocked over and false positives were recorded as the vehicles passed by in close proximity with the narrow work zone. At 0 feet away from the unit, the Intellicone alarm sound is 100 decibels and drops dramatically to around 70 dB at 50 feet. From around 70 decibels at 50 feet to approximately 55 dB at 300 feet, there is no discernible difference in alarm decibel levels. The transmission rate of radio waves between the sensor and PSA is effective up to 100% at a distance of up to 350 feet, while it is 0% at 450 feet. When a vehicle impacted the sensors at 30 mph, the workers’ maximum median reaction time was roughly 1.22 seconds in response to the alarm, which had a mean duration of 20 seconds. The AWARE system had a detection range up to 500 feet and the main alarm is triggered within a range of 300 feet from the main transmitter. Worktrax (the personal body alarm device) reliably produces vibratory and audio notifications. The AWARE’s internal stopping sight distance computations are precise enough to predict the alert’s timing. As per MnDOT observations during the AWARE testing, it was found that the system components were easy to deploy and can be effectively utilized in maintenance operations. The best-case use suggested by Caltrans professionals interviewed was that the WAS, along with the Intellicone and SonoBlaster, is useful in a ramp closure placed between barricades which has low exposure of the worker. The SonoBlaster is recommended by VDOT engineers to be used with an AFAD to maximize its effectiveness to prevent a work zone intrusion. The effective use of a SonoBlaster is in long-term work zones as the issues were considered priori to construction and the working personnel can be better prepared with solutions. The CO2 cartridges should be tested for the functioning, and if not operational, the crew should have extra CO2 cartridges available as a stand by. The AWARE system is suggested by the MnDOT for work zone maintenance operations. 3.3.3.6: Wearable Lights Wearable lights are worn by individual workers to enhance their visibility to oncoming drivers during nighttime work. Lights are available that can be worn on a hardhat, vest, or around a

76 person’s arm or leg. The lights are battery powered, can emit one or multiple different colors (e.g., white, red, amber), and often can be set to either static or different types of flashing modes. A variety of different sizes and styles are currently available for purchase at a relatively low cost. The Halo Light™ by Illumagear (https://illumagear.com/halo-sl/) is an example of one type of wearable light. Studies have been conducted to evaluate the impact and performance of the Halo Light™, and as a result, it is the focus of this case study. The Halo Light™ is a personal lighting safety system that attaches to most hard hats. The light illuminates the task area around the worker for a distance of up to 50 feet for a full 360˚ around the worker. The light can be seen by workers, equipment operators, and drivers in passing vehicles in all directions at all times, especially in low-light and darkened conditions. The Halo Light™ is waterproof, dust proof, and durable. The Halo Light™ costs approximately $130.00, is available in different variants such as Halo SL White, Halo SL Amber Brake, and Halo SL Red Brake, and can be purchased as per the user needs. The light comes with a rechargeable battery and battery charger. The dimensions of the Halo Light™ SL are: 262 mm wide, 332 mm long, and 28.5 mm thick. The specifications of the Halo Light™ SL are provided in Table 3.16. Table 3.16. Halo Light™ SL Specifications (https://illumagear.com/wp-content/uploads/2020 /05/Specifications-Sheet.pdf) Characteristic Specification Max. Light 360° mode: 276 lumens Spot Light 406 lumens Task Light 262 lumens Halo Light Weight 10 ounces Battery Weight 1.6 ounces Battery Type Single Lithium Ion 18650 cell; 3.5Ah Battery Life 1.5 - 121+ hours Water Resistant Submersion up to 1 m for 30 min Dielectric Strength 30,000V (min) International Protection (IP) Rating IP 67, Dust proof & Water Resistant Illumination Modes : Halo Mode Hi-Alert Mode Task Mode Dim Mode 276 lumens, 25.64 foot-candle Fluctuating luminosity [The light keeps revolving in circular pattern with different intensity of the light] 259 lumens, 82.4 candela 49 lumens Power 7.2 volts @ 0.6 amps Dielectric strength 9000V minimum Safety Standards UL24, UL1638, UL 8750, IEC/EN 60598-1, IEC/EN 60598-2-4, IEC/EN 62031, CSA9, CSA C22.2 No.9, CSA C22.2 No.250.13, Federal Communications Commission (FCC) Part15-b

77 The Halo Light™ can be set to different illumination modes that distribute the light in different patterns. The modes are as follows (Purdue ECT Team, 2015): • Halo mode: 360o ring visible over ¼ mile away in all directions • Hi-alert mode: Attention-getting rotating/pulsing light • Task mode: 200% power to front, fully flooding local task area • Dim mode: low power 360o ring of light Wearable lights are intended for use at night and can be worn in almost any work zone operation. With respect to work zone intrusions, the lights are intended to make drivers aware of the presence of the workers in the instance when an intrusion occurs. Some wearable lights have different settings for different situations. For example, for the Halo Light™, halo mode emits the full brightness of the light for 360-degree which helps the worker to see and move around and perform a number of jobs around the site. Hi-alert mode provides a unique visual identifier to maximize the chance of the user being seen through the use of the flashing lights in a 360-degree spin around the light. This mode is useful for flagging operations in order to be visible to passing motorists and construction equipment operators in the work zones. For the task mode, the front third of the Halo Light™ illuminates with maximum brightness to light up the work space or task area in a darkened area. Lastly, dim mode provides a 360-degree ring of light where the user is in close proximity with other workers having a face-to-face conversation. The dim mode is less bright and easier to look at from a close distance than the brighter modes. Virginia DOT employees interviewed for the case study indicated that Halo Lights™ are used to ensure maximum visibility of the workers to the road users during harsh conditions such as a flooding and snowing weather. VDOT employees also mentioned that Halo Lights™ are is used on nighttime projects and work zones with low-light conditions or darkened areas. The interviewees felt that the Halo Light™ features and specifications match with any type of work zone activity. The VDOT employees interviewed expressed that each worker was given a duty bag for use on a work zone. The bag contained a Halo Light™, first-aid kit, flashlight, light arm bands, and additional items. The project staff received positive responses from the workers who received the duty bag. The workers did not have any issues with the gear in the duty bag as it was easy to operate and use. The VDOT workers were happy with the performance of the Halo Light™ and, when worn, the workers felt they were more visible in hazardous conditions. The Halo Light™ could be used in hazard situations occurring either in U.S. or other worldwide locations, and adapts to the conditions at hand. In one example, the light was used during a tree removal activity in a work zone, wherein the Halo Light™ made sure the workers were visible to the passing drivers who were far away. Some workers felt that the Halo Light™ is heavier than anticipated and cumbersome to wear as it does not fit on different hardhat styles, such as the task hardhat style. Changes in weather conditions, such as the presence of fog, mist, and smoky air area contribute to poor visibility of the Halo Light™ and minimizes its visibility. Use during natural disasters such as flooding, snowfall, and heavy rainfall reduces the effectiveness of the Halo Light™ as it is

78 susceptible to wear and tear at a faster rate than specified. When the workers were soaked and saturated by heavy rainfall, the Halo Light™ was not able to withstand the wet weather and gave way to moisture or water. Charging of the Halo Light™ batteries was seen as an issue. The Halo Light™ stays active while the battery is charged and taken care of, and the battery life weakens when not charged or used for a long time. Previous versions of Halo Light™ used a disposal battery, and the battery issue is rectified in the latest versions using rechargeable lithium ion batteries. Suggestion from VDOT employees included improved battery life, a ring that is more streamlined, and adaptability to fit on different styles of hardhats. VDOT employees felt that use of the Halo Light™ is not compatible when workers work during a storm to put fuel in different vehicles, and running a chainsaw to remove tree debris. The VDOT staff recommended the workers use other sources of light, such as a light plant. The Halo Light™ is a new technology and the VDOT employees interviewed suggest other DOTs try out new safety technologies rather than using previous technologies or traditional methods. The important highlights are that the Halo Light™ functions as specified, it is compatible with most types of work zone activities, and ensures the employee are visible to the road users and other workers. 3.3.4. WZIT Case Study Analysis The case studies provide unique insight into current use and practical application of WZITs, along with the barriers encountered and enabling practices in their implementation. Adopting and implementing WZITs is an important decision as DOTs strive to improve work zone safety yet are limited in the resources that can be applied to their adoption and comprehensive implementation. Some DOTs are willing to acquire and try WZITs that have not yet been shown to provide long- term benefits and may still need some continued development to make their implementation effective. The promise that some technologies provide in terms of improving safety and preventing injuries and fatalities is often attractive in the decision-making process and drives the decision to acquire and use the technology. To further understand the key considerations associated with WZIT implementation, the research team performed content analyses of the case studies. The analyses were intended to expose trends in DOT thinking along with barriers and enablers of WZIT implementation. The analyses reveal the following commonalities amongst the applications of the WZITs and the decisions that went into their adoption and implementation: • Presentation, clarity, and accuracy for the passing motorist are critical to acceptance. Significant concern was given to how drivers recognize, interpret, and react to messages and alerts created by WZITs. The technologies must provide clear messages that are easily understood and direct drivers to safe passage through the work zone. Limited value was given to technologies that required significant attention to ensure clear communication with drivers, such as PCMS signs that needed extensive attention when orienting and locating the devices on the roadway, technologies that emit false positive or false negative alarms,

79 and variable messaging information that is not real-time and does not reflect actual work zone conditions. • The ability to easily transport and move the technology into position is also a concern. Technologies that are difficult to mobilize to the site, place on the roadway, and orient for optimal application are valued less and, over time, their use will diminish. Ease of use is highly valued, especially if minimal additional resources are required to accompany their implementation. For mobile operations, those technologies that can easily be attached to the equipment and move along with the operation, rather than require additional support equipment and resources, are preferred and ultimately more effective because they are used regularly. • Most of the WZITs available are applicable to the wide variety of types of work performed on roadway construction projects and during maintenance activities. The WZITs can be used during both daytime and nighttime operations, in urban and rural settings, and on roadways of all types. In some cases, the technologies, such as a mobile barrier, are more applicable to small work areas due to their size. Having the ability to extend their coverage to a larger work area, like the Intellicone, is a benefit. • When possible, positive protection to prevent an intrusion is desired. A mobile barrier is a good example of a WZIT that provides positive protection. Relying on warnings/alerts, variable messaging, and driver behavior to prevent and/or mitigate intrusions has less value. The effectiveness and reliability associated with the type of control are the basis of the hierarchy of controls for hazard mitigation. The “hierarchy of controls” or “order of precedence” is well known by safety and health professionals as a guide to follow to provide and manage a safe work environment. The hierarchy was initially conceived to reinforce the belief that controlling the hazard at the point closest to the source of the hazard as possible is the best approach. The hierarchy is a list of actions for reducing risk of injury that is ordered based on the reliability and effectiveness of each action. The order of precedence can be stated as follows, with the items listed from 1 to 5 in order of decreasing priority, reliability, and effectiveness: (1) eliminate the hazard; (2) utilize a different material, process, or product to reduce the hazard; (3) provide engineering controls to protect workers from the hazard; (4) warn workers of the hazard and train them how to act safely; and (5) provide personal protective equipment. Those WZITs that perform higher on the hierarchy, i.e., eliminate the hazard, will be more effective in their ability to prevent work zone intrusions. • The presence of supporting capabilities amplifies the WZIT’s value. For example, a mobile barrier may have work area lighting attached to the trailer which illuminates the work for the workers. Additional capabilities that support the work operation can eliminate equipment that may create a hazard and minimize exposure to workers. • Cost is a limiting factor when the initial purchase and long-term maintenance costs are high. Cost-benefit analyses should be conducted to confirm a positive return on investment. • Presently, those systems that rely on artificial intelligence technologies are limited in their ability. In some cases, the presence of false positives/negatives inhibits adoption of the technology and further research and development of the technology are needed. The acceptable level of automation for such technologies is a concern. Levels of automation

80 describe the ability of a technology to perform operations normally performed by a human, from initial data gathering through analysis of the data, identification of a risk, determination of alternatives to mitigate the risk, selection of an alternative, and finally implementation of the selected alternative. A technology with a higher level of automation can perform more tasks and decisions. However, doing so requires intelligent systems. Systems that utilize artificial intelligence have not yet been developed to the point where confidence in their reliability is high. In addition, for some work operations, it may not be desirable to rely on technologies to perform some tasks and decisions, especially those that are complex, critical, and greatly affect motorist and worker safety. Careful consideration should be given to the level of automation of the selected WZIT. • Technology readiness varies amongst the different categories/classes of WZITs, and even amongst the WZITs within a single category/class. As a result, those DOTs that are more risk averse in their adoption of new technologies may be limited in the suite of technologies that can be selected. 3.4 Decision Support Tool To ensure translation of research to practice, the findings from the literature review and the case study interviews conducted in this project was implemented into a DSS for ease of use and access by DOT employees and contractors. The DSS was implemented as a web-based system to enable ready access to interested personnel without the need for downloading and installing special software. This format also ensures that policies regarding installation of third-party software on organization (e.g., DOT) computers are not violated, and thus makes the tool widely accessible. Figure 3.7 shows the framework and workflow for DSS implementation and deployment. Figure 3.7. Framework and workflow for DSS implementation The following subsections describe the development of the database and DSS in more detail. Case Studies WZIT Database (Backend) Website with user interaction Survey Results Literature Review

81 3.4.1 Synthesis of Database The first step towards developing the DSS involved creating a database that summarizes and synthesizes information that influences the decision to use a technology. Information on the utility of the 15 WZITs evaluated in this study was extracted from multiple sources. First, the researchers extracted key WZIT implementation factors from the WZIT taxonomy created. These factors are associated with the project/roadway characteristics (work zone duration, location of work activity, type of work zone, time of day, and motorist speed, for instance), technology characteristics (functionality, target of intrusion alert, connectivity, and output features, for instance), and implementation and operational factors (cost, technology maturity, maintenance, and usability, for instance). Subsequently, the researchers assessed each WZIT using these factors, adopting a four- point scale (0 = NA; 1 = low impact/effectiveness/applicability; 2 = moderate impact/effectiveness/applicability; 3 = high impact/effectiveness/applicability). Points were assigned to each technology and factor based on information extracted from archival data, reports, questionnaire surveys, and interviews. Although somewhat subjective, this process of assigning points relied on mostly quantitative or objective information. Each member of the research team reviewed the synthesized databases to verify the accuracy and reliability of the point allocation. 3.4.2 Development and Use of DSS The DSS was created using Google’s Webapps feature. This format was used in order to be able to utilize Google Sheets as the database to store information about technologies and its ratings. Using Google Sheets enables ready access to be shared with multiple stakeholders and an intuitive interface to update the backend ratings if needed. The link to the webpage containing the DSS is available from the authors to interested parties upon request. The DSS webpage consists of two sections – the selection criteria section and the applicable technologies section. The webpage containing both sections is shown in Figure 3.8. The selection criteria section consists of a group of selection drop boxes that enable the user to specify various constraints or conditions that they would like the technology to meet. The available constraints/criteria correspond to the database headers and currently enable the user to select a variety of options related to roadway characteristics, technology characteristics, and operational features of technology. The list of WZIT technologies that is displayed changes dynamically in response to user input and in accordance with the ratings provided in the database. As an example, when the warning for only workers and the V2I connectivity constraints are selected, the list is shortened and its order is changed, as seen in Figure 3.9. The list of eligible technologies is shortlisted from all available technologies by first checking user input against the database to see if a specific WZIT technology has a non-zero score for all the user- selected criteria. If that is the case, that technology is given a score that is calculated by summing up the non-zero scores for the user-selected criteria. This score is then used to rank the technologies from highest to least applicable for all of the user’s constraints. Additional work for the DSS can involve providing more resources relating to the technology such as manufacturer links and user reports. It is anticipated that this DSS can guide and inform users on what technologies apply to which scenarios on the jobsite. This DSS also succinctly captures the major findings of the research synthesis project and provides it to stakeholders in a very accessible format.

82 Figure 3.8. Screenshot of WZIT DSS

83 Figure 3.9. DSS showing dynamic changes in response to user input

84 3.5 WZIT Guide The results and insights from this study’s literature review, survey, and case studies were implemented into a guide for practical use by DOTs and roadway contractors. The guide is available from NCHRP in a separate, standalone document. This section provides an overview of the content and organization of the guide. Chapter 1 of the guide provides an introduction to the document and describes its purpose and intended audience. Chapter 2 provides the background of the safety concern of work zone intrusions by discussing the various aspects of work zone intrusions from an analysis of previous incidents. A review of research literature and commercially available products was conducted to assemble a list of representative technologies. Fifteen such technologies were identified during the study, and these have been presented in the guide in terms of the following aspects: (1) technology description; (2) current and potential work zone applications; (3) effectiveness; and (4) implementation guidelines. The 15 technologies are presented in Chapter 3 of the guide. The technologies are classified based on their major functionality into one of three groups: Positive Protection Systems, Worker Warning Systems, and Driver Warning Systems. The guide also presents information about the use of the web-based DSS wherein users can select from various project, technology, and implementation constraints and view a list of applicable technologies that are arranged in order of suitability. A web-based implementation provides ready and widespread access to this information and enables uploading the information with newer technologies with time. A description of the DSS is provided in Chapter 4 of the guide. Thus, the guide provides information regarding the cause of work zone intrusions, describes potential technological solutions, and describes a DSS to disseminate information about available technologies. It is anticipated that this information can enable highway construction and maintenance stakeholders to objectively select appropriate technologies for their projects to prevent and mitigate the impact intrusions on their work zone. 3.6 Evaluation of Research Products As indicated in Chapter 2, the research team utilized a select group of practitioners (SMEs) to provide practical review of the study resources and outputs during the course of the study. Review and input related to the DSS and WZIT guide are beneficial to their applicability and clarity. To assist with the review, the research team developed an assessment tool/survey to guide reviewers through the assessment process. The assessment survey asked participants to assess the research products using a 1 to 9 scale (where 1 = very poor/totally disagree; 5 = average/neutral, and 9 = excellent/totally), based on the following evaluation criteria: a. Comprehensiveness b. Ease of use c. Ease of understanding d. Practicality e. Adaptability f. Engagement

85 g. Effectiveness The products and assessment survey were sent to the three SMEs who reviewed previous research documents. The research team also sent the products and assessment survey to three additional DOT employees to review and provide feedback. By involving additional reviewers, the research team ensured that essential external perspectives were considered in the review and feedback phase. Input was received from a total of three reviewers. All of the reviewers have more than 10 years of experience working with work zone traffic control. Their roles include managing work zone traffic control operations and work zone safety management. One of the reviewer’s roles is in technology adoption and implementation, while another reviewer is involved in work zone technology research. The rating results are shown in Tables 3.17 and 3.18. The results suggest that the resources are relatively easy to understand. The WZIT guide was found to be highly engaging and comprehensive, and also above average in terms of its effectiveness. Practicality, ease of use, and adaptability of the guide to multiple types of projects were rated lower, but still rated as average to above average. Like the guide, the DSS was also rated highly in terms of ease of understanding. Above average ratings were given for ease of use, engagement, effectiveness, and comprehensiveness of the DSS. A less than average rating was given for the practicality of the DSS. Table 3.17. WZIT guide assessment ratings Assessment Statement Mean Median SD The information in the guide is easy to understand. 7.3 7.0 0.6 The guide is practical (i.e., provides accurate and consistent information). 5.7 6.0 1.5 The guide would be easy to use on roadway projects. 5.7 6.0 1.5 The guide is adaptable to multiple types of projects. 6.3 5.0 2.3 The guide is engaging. 8.3 9.0 1.2 The guide is effective (sufficient in breadth and depth). 7.0 7.0 2.0 The guide is comprehensive. 8.0 9.0 1.7 Table 3.18. DSS assessment ratings Assessment Statement Mean Median SD The information in the DSS is easy to understand. 7.7 7.0 1.2 The DSS is practical (i.e., provides relatively accurate information). 4.7 4.0 1.2 The DSS would be easy to use on roadway projects. 6.0 6.0 1.0 The DSS is adaptable to multiple types of projects. 6.3 5.0 2.3 The DSS is engaging. 6.7 6.0 2.1 The DSS is effective (sufficient in breadth and depth). 6.3 6.0 0.6 The DSS is comprehensive. 6.7 6.0 2.1 All of the reviewers provided general comments regarding their impressions of the WZIT guide and DSS. A summary of the comments is provided below:

86 • Guide: o Need to also acknowledge that work zone crashes can also result from human error (traveling public and workers). o Use of an AFAD for 24-hour periods in rural settings should be specified more often and included as a bid item. o Some of the technologies are not commercially available, so I don’t know how the guide can be practical if some of the solutions are not available. o The guide is easy to use for a list of available solutions that could be used, and is a good step toward trying to encompass all of the different types of work zones, but I think it still falls short in being a tool to use to find a specific solution for a specific work zone. Engineering judgement still needs to play a part. o The guide is effective at education of all of the different technologies currently available and effective at providing the user available solutions for a certain “type” of work zone. • DSS: o Most of the DSS is easy to understand. Some classifications may need to be clearer, such as connectivity type. o Some of the technologies are not commercially available, so I don’t know how the DSS can be practical if some of the solutions are not available. o Easy to get a list of available technologies. o The tool wasn’t intuitive to guide me through the process to select different road types and conditions. Some simple instructions at the top to include something like “Select Conditions below and the results with automatically refresh on the right” would be helpful. o Sometimes I might not know what I want for technology on a project for intrusion alerts, so that was left “all” for my first search. I don’t know why I might pre-select a technology, but there you go. I rated the lowest for this tool on practicality. Once I looked through my project choices and saw the items on the right, I liked how it narrowed choices based on my input, but I have no other information than what you provided. For example, if I didn’t know about mobile barrier, there would be no further way for me to find out unless I left this tool and searched on Google. I understand you don’t want to endorse one piece of equipment, but pictures and/or links to more information seems like it is needed with this tool to be more practical. o Some categories were general. For example, intrusion detection with Bluetooth. Is there a product on the market for this setup or do I have to create it on my own? o A category for road geometry was not listed. Sometimes I have heavy traffic in good weather conditions, but there was no way for me to indicate a twisted road, or has a steep grade. Now I’m not sure technologies are the solution to that – but isn’t one answer ALWAYS to redesign the traffic control plan and give more space/warning/distance? I would argue that the design aspect seems a bit lost here with this tool. Although design is not a ‘technology’ for intrusion alerting, it seems it should be on the forefront of every contractor’s mind. Is there a best practices

87 design standard that can be added as a choice for this tool? That’s probably out of scope here, but something to consider. The WZIT guide and DSS were modified to incorporate the input received from the SMEs.