Use of Smart Work Zone Technologies for Improving Work Zone Safety (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.

18 Use of Smart Work Zone Technologies for Improving Work Zone Safety initial criteria to assess the need for smart work zone technologies (based on user delay and traffic volumes), an overview of several smart work zone technologies, a design process checklist, and guidance for system setup, procurement, and maintenance. • A smart work zones guide for the Connecticut DOT provides an overview of smart work zone technologies, information on the Connecticut DOT’s policies for smart work zones, typical applications (Figure 12), and guidelines for project feasibility assessment and implementation of smart work zone technologies (IBI Group 2017). DOT Implementation Tools for Smart Work Zone Technologies • Decision-making tools developed by the Minnesota DOT include a smart work zone toolbox and scoping decision tree. The smart work zone toolbox furnishes guidance on selecting appropriate smart work zone technologies (Minnesota DOT 2020d). The toolbox divides the technologies into three categories (traffic, vehicle, and environmental) and includes a general overview of common components and information such as warrants, benefits, options, and layout drawings for 13 smart work zone technologies. The scoping decision tree (Minnesota DOT 2019) facilitates the process of determining what level and type of smart work zone technology is warranted on a given project. The decision tree divides technologies into four categories (traveler information, motorist advisory, motorist warning, and route manage- ment) and provides information regarding benefits, costs, and other considerations for various smart work zone technologies. • An operations bulletin from the New Jersey DOT includes a chart with scoring criteria to help assess if a Real Time Work Zone Travel System (RTWZTS) is warranted on a given construc- tion project (New Jersey DOT 2013). The chart considers 10 factors, such as annual average daily traffic (AADT), volume per lane, and availability of alternative routes. • A smart work zone toolbox from the New Hampshire DOT furnishes guidance on selecting appropriate smart work zone technologies and divides the technologies into three catego- ries: traffic, vehicle, and environmental (New Hampshire DOT 2011). Summary sheets with information regarding conditions suitable for deployment, benefits, layout, options, and other notes are provided for 12 smart work zone technologies. Additional implementation guidance from the New Hampshire DOT includes a step-by-step process for implementing smart work zone technologies, a form with scoring criteria to determine if smart work zone technologies would be beneficial on a given project based on five factors, and a table showing smart work zone systems mapped to the traffic characteristics they are capable of addressing (New Hampshire DOT 2016). • A scoring chart developed by the Massachusetts DOT allows for an evaluation to determine if deployment of smart work zone technologies at a work zone has the potential to be benefi- cial (Massachusetts DOT 2016b). The evaluation is based on seven factors, including roadway geometry, work zone duration, queuing, delay time, and impacts for commercial motor vehicles. • Smart work zone guidelines from the Texas DOT provides direction for the design and deployment of several smart work zone technologies, including queue detection, speed moni- toring, construction enter or exit notifications, travel time estimation, incident management, and overheight warnings (Texas DOT 2018). An appendix to the guidelines includes example cost information for various smart work zone technologies. In addition to the guidelines, the Texas DOT also developed a go/no-go spreadsheet tool (Texas DOT 2020), which includes scoring charts (Figure 13) and a decision tree methodology to determine if a given smart work zone technology should be implemented on a project. • A smart work zone design tool was developed in a research study sponsored by the Arizona DOT (Kimley-Horn and Associates 2020). The tool, which is spreadsheet-based, asks the designer for input on five factors and other parameters. Based on the input provided, the tool provides information regarding the types of smart technologies that should be deployed and their placement within the work zone (Figure 14).

Literature Review 19   Source: IBI Group 2017. Figure 12. Typical application of real-time traveler information system for Connecticut DOT (CCTV = closed circuit television; CTDOT = Connecticut Department of Transportation; MUTCD = Manual on Uniform Traffic Control Devices; PVMS = portable variable message sign; SWZ = smart work zone).

20 Use of Smart Work Zone Technologies for Improving Work Zone Safety Source: Texas DOT 2018. Figure 13. Scoring sheet for temporary construction equipment alert system from Texas DOT’s go/no-go spreadsheet tool.

Source: Kimley-Horn and Associates 2020. Figure 14. Example output from Arizona DOT’s smart work zone design tool.

22 Use of Smart Work Zone Technologies for Improving Work Zone Safety Evaluation Studies, Guidance, and Standards for Specific Smart Work Zone Technologies In addition to general guidance for smart work zone technologies, guidance and evaluation studies are available for specific types of smart work zone technologies. These resources are described in the following sections, which are organized by type of smart work zone technology. Traveler Information Systems Evaluation Studies for Traveler Information Systems Several research studies have shown PCMSs to be effective in reducing vehicle speeds. For example, results from a field study of a work zone in Kansas showed that the use of graphic- aided PCMSs to notify drivers about a work zone and flagger ahead led to reductions in vehicle speeds of 13% to 17% (Huang and Bai 2014). A field evaluation of a PCMS on rural highways operating with one-lane, two-way work zones in Kansas and displaying the message “Road Work Ahead” found that vehicle speeds were reduced by almost 5 mph with the PCMS switched on (Li et al. 2010). An assessment of dynamic message signs (DMSs) was undertaken in Missouri using motorist surveys, field evaluation, and a simulation study (Edara et al. 2011). The survey results indicated that motorists’ opinion of DMSs was highly favorable. Findings from the field study on two I-55 work zones showed that average vehicle speeds decreased by 3.64 mph and 1.25 mph with the DMSs displaying information regarding the presence of the work zone. The simulation study, which investigated the effects of DMSs on traffic diversion for a bridge closure (Figure 15), estimated monetary benefits of DMSs ranging from approximately $5,000 to $56,000. Researchers have also investigated the effects of PCMS location and message content on vehicle speeds. An analysis of vehicle speed profiles using data from two field experiments in Kansas found that location for optimal deployment of an upstream PCMS notifying drivers about a work zone and flagger ahead is located 556 to 575 feet from the W20-1 sign (Li and Bai 2012). Results from a simulator study to assess effects of four PCMS messages on vehicle speeds indicated that the message sign “Prepare to Stop/Stopped Traffic Ahead” resulted in the greatest speed reduction immediately upstream of the lane closure (Bham and Leu 2018). In addition, drivers rated the message “Speed Ahead 30 mph/X Min to End of WZ” the highest for effective- ness in a post-simulator survey. Various studies have shown smartphone-based alert systems for work zones to be effective in encouraging safer behavior by drivers. A driving simulator study of three types of in-vehicle Source: Edara et al. 2011. Figure 15. Example use of DMS during bridge closure on I-55 in Missouri.

Literature Review 23   messaging interfaces (PCMS, smartphone with auditory messages only, and smartphone with audio-visual messages) (Figure 16) determined that driving performance based on metrics such as speed and lane deviation improved for smartphone messages as compared to the PCMS infor- mation (Craig et al. 2017). Another driving simulator study tested four different types of messages (visual, sound, male voice, and female voice) for a smartphone-based warning system for work zones, with the results indicating that voice warnings led to earlier deceleration, safer speed and deceleration rate, and longer headway time and distance (Rahman et al. 2016). An Android smartphone app which provides warning messages in work zone termination areas was devel- oped and evaluated using a driving simulator and participant survey (You et al. 2016). The messages, particularly the female voice and beep tone versions, significantly improved driver behaviors by decreasing speed and brake reaction distance. A field assessment of another smart- phone app developed by Liao (2019) found that the app is capable of receiving and announcing messages from Bluetooth low energy (BLE) beacons at an average distance of 127 meters. Other research studies have assessed the use of other technologies such as connected vehicles and radio frequency identification (RFID) to communicate information about work zones to drivers. A work zone driving simulator experiment with professional drivers found that the use of the Wyoming Connected Vehicle Pilot’s Transportation Traveler Information Messages on rural I-80 resulted in lower crash risk, as indicated by higher time-to-collision and lower decel- eration rate (Yang et al. 2020). An investigation of work zone countermeasures with the potential for implementation using vehicle-to-infrastructure (V2I) and infrastructure-to-vehicle com- munication found that drivers who were distracted or under the influence of alcohol were over- represented in work zone crashes and that signs outside of a vehicle showed limited potential to alter these drivers’ behaviors (Misra et al. 2018). Based on these findings, the researchers suggested that V2I systems might be more effective than vehicle-to-vehicle communications by allowing DOTs to send customized, in-vehicle messages. A driver smart assistance system (DSAS) that uses RFID to send verbal and visual messages (“Road Work Ahead,” “Speed Limit 30 mph,” and “Right Lane Closed”) to drivers as they approach a work zone was developed by Qiao et al. (2014). The DSAS was evaluated on a test track in Houston, and the results indicated earlier deceleration by drivers. Guidance and Standards for Traveler Information Systems Various forms of national and DOT guidance on the use of traveler information systems, espe- cially PCMSs, are available. A guidance document for traveler information systems, including an overview of different methods (Table 2), safety considerations for PCMSs, and DOT examples, Source: Craig et al. 2017. Figure 16. Example message types evaluated in driving simulator study.

24 Use of Smart Work Zone Technologies for Improving Work Zone Safety is available from the Roadway Safety Consortium (2011). The document provides suggestions for improving effectiveness of traveler information systems, such as the use of social media and the inclusion of PCMS guidance in DOT standard drawings. The Manual on Uniform Traffic Control Devices (MUTCD) provides some guidance on CMSs (Section 2L) and PCMSs (Sec- tion 6F.60) (FHWA 2009). At the DOT level, customized guidance on the use of PCMSs for the Wisconsin DOT was developed by Paulus (2016). The guidance provides information such as typical duration and phase length of the message display, PCMS placement, and sample messages for different scenarios for advanced closure notices and advanced warnings. A handbook from the Oregon DOT dis- seminates basic guidance on the placement (Figure 17), delineation, and use of PCMSs (King and Strategies Benefits Implementation Factors to Consider Dynamic Message Signs/Portable Changeable Message Signs • Can provide real-time information directly to driver while en route • Highly valued by motorists • Amount of information that can be disseminated is limited • Proper message design is critical Highway Advisory Radio • Can provide real-time information directly to driver while en route • Can provide more information than is possible with DMSs/PCMSs • Proper message design is critical • Requires motorists to seek out information Media Alerts • Easy to distribute • Low-cost activity • May only reach local travelers • Agency does not have total control over information presented 511 Traveler Information Telephone • Allows information to be accessed whenever it is needed • Can allow feedback by travelers to be recorded if desired • Audience must be informed of the system’s existence • System is not available in all areas Traffic Updates to GPS Navigational Devices • Can provide information directly to driver • Services is not available in all areas • Will reach only those travelers who subscribe to service Project Website • Easy way to provide current conditions to those planning travel • Mobile access via smartphone applications • May not reach entire target audience (i.e., those without access to the Internet) Real-Time Traffic Information Through Public Website (e.g., Google Maps) • Easy to distribute • Low cost • May not be accessed by entire target audience • Warning not to access while driving should be included Email Alerts • Very low cost • Easy to distribute • Requires audiences to sign up to receive information, or for the agency to seek out email addresses • Criteria must be established as to when alerts will be sent Text Messages • Very low cost • Easy to distribute • Amount of information that can be disseminated is limited Social Media • Low cost • Easy to distribute • May not reach entire target audiences (i.e., those without access to the Internet or social media) Table 2. Summary of methods to disseminate traveler information.

Literature Review 25   McCrea 2018). The handbook includes a list of standardized messages for various work zone activities and other events. Guidelines for use of PCMSs from the Michigan DOT describe the following high priority uses: advance notification of ramp closures, lane closures, roadway closures, and planned maintenance work (Michigan DOT 2011). Some DOTs include additional require- ments for PCMSs in their DOT MUTCD or MUTCD supplement. For example, the Maryland State Highway Administration requires PCMSs to be visible from at least 0.5 mile and legible from a minimum of 900 feet under both daytime and nighttime conditions for state-owned roads (Maryland State Highway Administration 2011). Queue Warning Systems Evaluation Studies for Queue Warning Systems Research studies on QWSs have investigated various operational aspects of these systems. Findings from an evaluation of a dynamic QWS based on microsimulation suggested 35 mph as the queue detection threshold, 55 mph as the threshold for a “Slow Traffic” message, and 5 minutes as the speed aggregation interval (Pesti et al. 2013). In addition, the use of half-mile detector spacing instead of one-mile detector spacing improved accuracy. Two queue warning and travel time estimation systems were used in a work zone on I-94 in Minnesota and evaluated for accuracy and latency (Petersen et al. 2014). The evaluation determined that both systems were operating within acceptable tolerances, and suggested modifications to the Minnesota DOT’s procurement specifications were offered. A queue alert system based on probe vehicle data was developed and tested at six work zones in Indiana (Mekker et al. 2017). When queues grow to be over a mile long, or speeds drop by more than 15 mph at the back of the queue, the system sends alerts to selected members of the Indiana DOT and allows them to prepare responses. The alerts were verified using still camera images, work schedules, and crash reports. Research studies have shown QWSs to be effective in reducing crashes and speed variance. Two queue detection systems using the intelligent lane control signals of an existing active traffic management (ATM) system were developed and evaluated at two highway work zones in Minneapolis, Minnesota (Figure 18) (Hourdos et al. 2017). The results indicated that crashes and near-crashes were reduced by 22% and 54% respectively with one system, while lower speed variance was observed with the second system. A study of the I-35 corridor in central Texas reported the following values of Crash Modification Factors (CMFs) for QWSs and por- table rumble strips deployed during nighttime lane closures: 0.717 (non-queuing conditions) and 0.468 (queuing conditions) (Ullman et al. 2018a). An economic analysis found that the use of the QWS on this corridor attained its break-even point after 95 to 190 nights of use based on crash reductions (Ullman et al. 2016). An evaluation of a QWS on I-43 in Wisconsin found that the implementation of the system reduced total crashes by 15% and injury crashes by 63% (Schulze 2018). Tennessee DOT indicates in public outreach materials for its Protect the Queue Source: King and McCrea 2018. Figure 17. Example of PCMS located outside driver’s cone of vision.

26 Use of Smart Work Zone Technologies for Improving Work Zone Safety program that secondary crashes decreased by 19% during the first 6 months of program deploy- ment compared to the same time period in the previous year (Tennessee DOT n.d.). Guidance and Standards for Queue Warning Systems A concept of operations document for the use of QWSs from the Minnesota DOT provides information regarding signage configuration (Figure 19), stakeholder needs, operational con- cepts from different perspectives, system components, and scenarios for operation (Minnesota DOT 2015a). High-level system requirements for QWSs from the Minnesota DOT state that the system shall be 95% accurate at monitoring vehicle speed, at any speed (Minnesota DOT 2018). Plan drawings for QWSs, including device layouts and message content for PCMSs, are available from the Texas DOT and the Washington State DOT (Texas DOT 2019, Washington State DOT 2021). Dynamic Lane Merge Evaluation Studies for Dynamic Lane Merge Research studies have shown that dynamic lane-merge systems can help to improve traffic operations and safety in work zones. An assessment of dynamic early merge and late merge using data from work zones on I-95 in Florida and microsimulation found that work zone capacity was significantly higher for early merge than for the Florida DOT’s standard maintenance of traffic (MOT) configuration (Harb et al. 2011, 2012). In public outreach material on dynamic lane- merge systems, the North Carolina DOT reported that the deployment of the system reduced the maximum queue length from 8 miles to 2 miles in a work zone on I-77 (North Carolina DOT 2019). Results from a driving simulator study of the impacts of traffic volume and dynamic merge messaging on merge location and throughput indicated that drivers merged sooner into the open lane under the early merge condition than under the late merge condition and that late merging led to higher throughput in high-volume conditions (Weaver at al. 2019). A benefit-cost analysis of the dynamic late-merge system found the system to be beneficial when the value of travel time savings exceeds $4.85/hour (Figure 20) (Datta et al. 2007). An evaluation of dynamic lane merge on US-69 in Overland Park, Kansas, found that there were three fewer crashes per Source: Hourdos et al. 2017. Figure 18. Warning messages displayed from QWS on I-35W in Minneapolis, Minnesota.

Literature Review 27   Source: Minnesota DOT 2015a. Figure 19. Typical layout of work zone signage for QWS from Minnesota DOT concept of operations.

28 Use of Smart Work Zone Technologies for Improving Work Zone Safety week as compared to another nearby work zone on the same corridor (Kansas DOT 2016a). The study estimated total costs of approximately $100,000 for 119 days of contracted work for the dynamic lane-merge system. Other research studies on dynamic lane-merge systems have investigated the effects of message displays and connected vehicles. A field evaluation of a portable dynamic lane-merge system compared two different message displays, with the display featuring an alternating roadwork graphic and speed limit sign on top and an alternating “Merge Left” message and merging traffic arrows on the bottom found to be the most effective at encouraging zipper merge behavior (Reinker et al. 2015). A dynamic lane-merge strategy based on the premise that all vehicles are fully automated, connected, and cooperative was developed and assessed using microsimulation (Ren et al. 2020). The results of the evaluation indicated that the system provided better operational and safety performance than popular merging strategies, such as late merge and early merge. Guidance and Standards for Dynamic Lane Merge Guidance regarding the use of dynamic lane-merging strategies is available from the American Traffic Safety Services Association (ATSSA) (2012). The ATSSA guidance includes a decision tree diagram to help practitioners determine when to use dynamic lane-merging systems (Figure 21) and sample system layouts. According to the ATSSA guidance, dynamic lane-merging strategies are more effective in work zones with traffic demands that vary. Work Zone Speed Control Technologies Various types of smart work zone technologies, including dynamic (variable) speed limits and speed display signs, can be used in an effort to control vehicle speeds in work zones. Evaluation studies and guidance documents for these technologies are described in the following sections. Dynamic (Variable) Speed Limits Evaluation studies of various types of dynamic (variable) speed limit systems have generally shown that the systems lead to reduced speeds, while existing DOT guidance on these systems is limited. An assessment of a portable variable speed limit (PVSL) system at four separate Source: Datta et al. 2007. Figure 20. Graph of benefit-cost (B/C) ratio versus value of travel time savings from Michigan DOT study on dynamic lane merge.

Literature Review 29   [Copyright © American Road and Transportation Builders Association (ARTBA)] Source: ATSSA 2012. Figure 21. Decision tree diagram for dynamic lane-merge strategies. deployments in Utah found that overall average speeds near the active work zone space were 15 to 25 mph lower than the original posted speed limit, and the system limited the length and duration of the speed reduction for motorists (Van Jura et al. 2018). Results from a field investigation of a variable speed limit (VSL) system in an Indiana work zone using vehicle-matching technology showed a maximum drop in mean speed of 4.7 mph, but three pairs of signs were needed to obtain meaningful speed reductions (Mekker et al. 2016). A study of the effectiveness of a variable advi- sory speed limit system using both field data in Missouri and microsimulation determined that average speeds were significantly lower and speed limit compliance was higher with the system (Edara et al. 2013b). However, in uncongested urban work zones, the speed variance was higher. A field assessment of a variable advisory speed limit system (Figure 22) at a long-term work zone in Utah concluded that the system was effective at reducing queues during weekend peak hours with traffic slowdowns but had no statistically significant effects during other time periods (Saito and Wilson 2011). VSL guidelines from the North Carolina DOT prescribe that VSL can only be used at a spot location in short-duration work zones where the existing speed limit is 65 mph or greater, and VSLs below 55 mph are highly discouraged (North Carolina DOT 2011). Display Signs, Warning Systems for Work Zone Speeds, and Automated Speed Enforcement Several research studies have found speed display signs to be effective in reducing vehicle speeds and deceleration rates in work zones, and specifications for these signs have been developed by DOTs. For example, a field evaluation of a system in Minnesota that displayed

30 Use of Smart Work Zone Technologies for Improving Work Zone Safety downstream speeds on PCMSs found that deceleration rates decreased when accurate infor- mation was shown to drivers (Hourdos et al. 2019). Results from an assessment of radar speed feedback signs (RSFSs) on multilane maintenance work zones in Oregon indicated that the use of the signs led to lower vehicle speeds and less speed variation between vehicles (Jafarnejad et al. 2017). Another field assessment of RSFSs in Arizona (Figure 23), which also included an alternating monetary fine message, determined that the use of the alternating messages led to a 50% reduction in the number of speeders driving 15 mph or more above the speed limit (Roberts and Smaglik 2014). Field testing of a speed-activated sign on two-lane highways in South Caro- lina showed an average reduction in mean speed of 3.3 mph on two-lane highways, with similar results on a multilane divided highway and Interstate freeway (Mattox et al. 2007). A field study in Kansas found that dynamic speed-feedback signs in two work zones led to significant reduc- tions in vehicle speed (Anderson et al. 2021). An analysis of commercially available speed data for connected vehicles showed that the use of presence lighting and digital speed limit trailers at a work zone on a rural Indiana Interstate resulted in median speeds decreasing by 4 to 13 mph between the hours of 11 p.m. and 7 a.m. (Sakhare et al. 2021). The Michigan DOT provides Source: Saito and Wilson 2011. Figure 22. Message sign for a variable advisory speed limit system in Utah DOT study. Source: Roberts and Smaglik 2014. Figure 23. Vehicle approaching speed feedback sign in Arizona study.

Literature Review 31   requirements for a temporary speed radar trailer in its special provisions (Michigan DOT 2021), while the South Dakota DOT lists conditions under which the use of RSFSs should be considered and placement guidelines for RSFSs in its construction manual (South Dakota DOT 2020). Other speed control strategies include speed warning systems and automated speed enforce- ment. A prototype speed warning system using smart barrels was developed and deployed at a highway work zone in Oklahoma (Zhang et al. 2012). The barrels include Doppler radars and radio nodes to monitor speeds and determine when a warning message (flashing LED lights) should be given to motorists (Figure 24). Significant speed reductions were noted during the prototype’s test deployment. The Pennsylvania DOT and Pennsylvania Turnpike Commission are partnering with the Pennsylvania State Police to implement an automated work zone speed enforcement program (Pennsylvania DOT 2020a). Work Zone Data Collection Technologies Results from a DOT survey conducted during a prior research study by Cheng et al. (2017) indicated that DOTs are at various levels of development for practices to manage work zone data. To help work zone stakeholders develop consistent data management approaches, a frame- work was developed by Stephens et al. (2019). The framework includes an organizational struc- ture for work zone activity data (WZAD) (Figure 25), WZAD dictionary, WZAD user needs and use cases, conceptual architecture and user interfaces for a work zone data system, and general implementation guidance. Various initiatives at the national and DOT levels seek to develop WZAD standards and improve access to WZAD. For example, FHWA is leading the Work Zone Data Initiative (WZDI), an effort which aims to establish standardized practices along with a data dictionary and imple- mentation guidance, for the management of seven types of WZAD (Figure 26), based on a stakeholder-driven and systems-driven approach (FHWA 2019a). The U.S. DOT is overseeing the development of the Work Zone Data Exchange (WZDx) Specification, which allows infra- structure owners and operators to provide WZAD for third parties with the goal of improving Source: Zhang et al. 2012. Figure 24. Layout of smart barrel system (represented as network nodes EN#1 through EN#10) for monitoring work zone speeds.

32 Use of Smart Work Zone Technologies for Improving Work Zone Safety Source: Stephens et al. 2019. Figure 25. Organizational structure for WZAD (TMC = traffic management center; ATMS = advanced traffic management system; RCRS = Road Condition Reporting System; ITS = intelligent transportation systems). Source: FHWA 2019a. Figure 26. Seven types of standardized WZAD being developed in WZDI. safety through universal access to WZAD (U.S. DOT 2021, GitHub 2021a). Release v3.0 of the WZDx is available on GitHub (2021a). Regarding DOT efforts in work zone data technologies, survey results from NCHRP Synthesis 561 indicated that at least 88% of responding DOTs use probe-based speed data, with some DOTs employing probe data for traveler information systems (Pack and Ivanov 2021). The Delaware DOT manages the Work Zone Incident Communication System (WZIC) project, which

Literature Review 33   includes the identification of data sets, data formats, and distribution procedures for commercial vehicles (CVs) (Delaware DOT 2020). The project uses a collaborative approach between the public and private sectors, with DOTs acquiring the data, and private companies maintaining and distributing the data. The Iowa DOT sponsored a research study to develop a research-grade work zone database that integrated planned 511 work zone data with data from smart arrow board deployments (Knickerbocker et al. 2020). Crowdsourcing data from applications such as Waze are being used to provide information to motorists and some DOTs. A research study by Amin-Naseri et al. (2018) evaluated the usefulness of crowdsourcing data from the Waze app as part of an incident management strategy by comparing 1 year of Waze data with incidents recorded in the advanced traffic manage- ment system (ATMS). Results indicated that Waze was able to identify incidents an average of nearly 10 minutes sooner than traditional approaches and provide information for 43.2% of ATMS crash and congestion reports. Through a partnership with Waze, the Wisconsin DOT obtains crowdsourcing traffic data from Waze and provides the official information from 511 to Waze (Wisconsin DOT 2017). The combined information is displayed on a website to enable better decision making by both motorists and the DOT. Research has also been conducted to investigate the feasibility of using crowdsourcing data as a screening tool to prioritize work zones for safety countermeasures. An analysis of com- mercially available hard-braking data for 23 Interstate work zones in Indiana found that there was approximately one crash for every 147 hard-braking events per mile (Desai et al. 2020). Hard-braking events and crashes by severity for a work zone on northbound I-65 are shown in Figure 27. (a) (b) (Copyright © Desai et al.) Source: Desai et al. 2020. N um be r o f H ar d- br ak in g ev en ts Figure 27. Hard-braking events and crashes by mile and severity for work zone on northbound I-65 for July 1, 2019, through August 31, 2019: (a) hard-braking events and (b) crashes by mile and severity [(i) = indicates that there were a higher number of hard- braking events and higher number of crashes near mile marker 61 on I-65 northbound.]

34 Use of Smart Work Zone Technologies for Improving Work Zone Safety Work Zone Location Technologies There has been some limited exploration and implementation of work zone location technol- ogies. A design prototype was developed for a Smart Work Zone Activity App (SWiZAPP) for collecting and reporting real-time information on work zone status (Adu-Gyamfi et al. 2019). The mobile application provides functionality for geolocation, mapping, image uploads, live updates of construction activities, and viewing real-time and historical data regarding work zone activity. In a study by Parikh et al. (2019), the Statewide Work Zone Information System (SWIS), a real-time database of active work zones that uses beacons placed on existing traffic control devices and interacts with various external sources of information, was created for the Minnesota DOT (Figure 28). The researchers indicated that additional field testing was required before SWIS could be implemented in Minnesota. Smart arrow boards are another type of emerging work zone location technology. Iowa is imple- menting smart arrow boards on all Interstate lane closures in 2021 and on all lane closures on the primary system in 2022 (Sprengeler 2020). The smart arrow boards generate data feeds with information on work zone location and start and end times that are compatible with the WZDI. The Iowa DOT developed specifications for the smart arrow boards with requirements for GPS and remote communications and a procedure for inspection and approval (Iowa DOT 2021a). The Minnesota DOT (MnDOT) performed some preliminary testing and test deployments of smart arrow boards and sponsored development of a document describing operational concepts, system requirements, and use scenarios for the smart arrow boards (Athey Creek Consultants 2018). A flowchart for the system, including interaction with other components, is shown in Figure 29. Work Zone Intrusion Alarm and Proximity Alert Systems Various systems for work zone intrusion alarms and proximity alerts have been developed, while the availability of guidance for these systems is limited. These evaluations and guidance documents are described in the following sections. Source: Parikh et al. 2019. Figure 28. Interaction between external data sources and SWIS.

Literature Review 35   Work Zone Intrusion Alarm Experimental studies of work zone intrusion alarm systems have shown mixed results, while the availability of existing guidance for these systems is limited. Results from pilot testing of three work zone intrusion alarm systems in California showed good performance for Worker Alert System, limitations with SonoBlaster, and successful trials for Intellicone after resolution of some issues (Khan et al. 2019). Supplemental plans for the Intellicone and Worker Alert System were developed as part of the California study. In a research study by Mollenhauer et al. (2019), a wearable system to alert workers to potential intrusions was developed. The system relies on data from connected and automated (CAV) vehicles to predict trajectories and send instructions to both drivers and workers. Evaluations of the system at a test track in Virginia showed that the ultrawide band sensors used did not have sufficient range for most work zones. During pilot testing of the SonoBlaster work zone intrusion alarm in New Jersey, concerns were noted about the system regarding setup, quality control, and reliability of the unit (Krupa 2010). A field study of mobile work zone alarms on a freeway work zone in Missouri indi- cated that most configurations increased the merging distances of vehicles when changing lanes (Brown et al. 2015). Guidance for system selection and system requirements for work zone intrusion alarms are available. Field experimental trials of the Intellicone and Traffic Guard work zone intrusion alert systems found that both systems performed satisfactorily based on several performance mea- sures such as sound levels, worker’s reaction time, and vehicle stopping distance (Marks et al. 2017). Based on the results of these trials and a field demonstration of the Advanced Warning and Risk Evasion (AWARE) system, a selection guide was developed for these systems (Table 3). Based on controlled testing, live work zone testing, and worker surveys in Tennessee, researchers suggested the use of Intellicone for long-term stationary work zones, AWARE Sentry for flag- ging operations, and the Worker Alert System for short-term and mobile work zones with no lane encroachment (Mishra et al. 2021). A document describing system requirements for work zone intrusion alarm systems is available from the Minnesota DOT (Minnesota DOT 2015b). Source: Athey Creek Consultants 2018. Figure 29. Flowchart showing interaction between Minnesota’s smart arrow board system and other systems (CARS = Condition Acquisition and Reporting System).

36 Use of Smart Work Zone Technologies for Improving Work Zone Safety Proximity Alert Systems Proximity alert systems for construction equipment and workers based on various technolo- gies have been developed and tested. The Embedded Safety Communication System, which uses tactile vibrations to communicate information about hazards to workers, was created by Park and Sakhakarmi (2019). Field trials to evaluate the system demonstrated that participants accurately and efficiently understood the communicated signals. In a research study by Cho et al. (2017), a proximity sensing system for construction equipment and workers was devel- oped using Bluetooth technology and compared with systems based on RFID and magnetic field. Results from field testing indicated that Bluetooth provided the best performance for cost, ease of installation, small-form factor, and flexibility in programming. The reliability of ultrasonic and pulsed-radar sensing technologies for the prevention of back-over accidents was assessed using four types of experiments, and the results identified strengths and limitations for the different technologies (Choe et al. 2014). A prototype system was developed based on the use of dedicated short range communications (DSRC) to convey visual guidance to construction vehicle operators regarding worker locations and to dynamically display appropriate speed limits and other warning messages on DSRC-equipped variable message signs (VMSs) based on the presence of workers (Figure 30) (Banaeiyan et al. 2016). Results from system testing indicated that worker location was accurate to within 6 feet and 20 degrees, and the VMSs were able to post appropriate speed limits based on workers’ presence. Notification of Construction Equipment Entering or Exiting The availability of direction for notification systems for construction equipment entering or exiting the work space is limited. Implementation considerations for these systems are discussed in a guidance document from the American Road and Transportation Builders Association (ARTBA) (ARTBA 2019). These considerations include using hardwired sensors or direct wire- less communication to ensure rapid generation of the alert and proper placement of the sensor so that it only detects construction vehicles entering or exit the work space. The Minnesota DOT developed a boilerplate special provision (Minnesota DOT 2020b) and typical applications for notification systems for construction vehicles exiting, crossing, and merging onto the highway (Minnesota DOT 2021b) (Figure 31). According to these typical applications, the PCMS is blank unless construction traffic is entering or exiting the traffic lanes. The Ohio DOT provides Situations Intellicone Traffic Guard Worker Alert System AWARE Longer than one day One day or shorter Mobile operation Taper longer than or equal to 1,500 ft Taper shorter than 1,500 ft × × × × × × × Table 3. Selection guide for work zone intrusion alarms.

Literature Review 37   Source: Banaeiyan et al. 2016. Source: Minnesota DOT 2021b. Figure 30. Conceptual layout for visual warning system based on DSRC. Figure 31. Minnesota DOT typical application for notification system for construction vehicles entering the highway with no acceleration lane. guidance and standards for these systems in its Traffic Engineering Manual (Ohio DOT 2021) and standard drawings (Ohio DOT 2018). Other Smart Work Zone Technologies Research studies have been conducted to investigate other smart work zone technologies, such as automated flagger assistance devices (AFADs), driveway assistance devices (DADs), and auditory warnings on roadside safety signs. Results from an evaluation of an AFAD with a

38 Use of Smart Work Zone Technologies for Improving Work Zone Safety CMS developed by the Missouri DOT (Figure 32) indicated that the AFAD reduced approach speeds, stopped vehicles farther back, and released trac faster than a human agger (Brown et al. 2018). A eld assessment in Michigan of DADs, which use a ashing red arrow to indicate when a motorist on a driveway or side street can enter the roadway aer yielding to trac in a work zone with one-way trac, found a compliance rate of 83%, with queues on side streets limited to three vehicles (Brookes 2016). A driving simulator study of auditory warning sounds used in conjunction with roadside safety signs showed improved driver compliance (Kang and Momtaz 2018). Other studies have evaluated technologies to help guide users through the work zone. A pro- totype smartphone app to help visually impaired pedestrians navigate through a work zone (Figure 33) was developed and evaluated for functionality, but the research results indicated that additional testing with users is required (Liao 2014). A eld study in Missouri found that the use of sequential warning lights in nighttime work zone tapers resulted in reduced speed, improved driver compliance, and a benet-cost ratio of 5 or 10 depending on how labor costs were calculated (Sun et al. 2011). Other research studies have investigated the use of a mobile work zone barrier system (Gambatese and Tymvios 2013) and a fully automatic device for cone placement and retrieval in work zones (Wang et al. 2019). Performance Measures for Smart Work Zone Technologies Guidance for Performance Measures for Smart Work Zone Technologies Guidance regarding the development and use of work zone performance measures is avail- able from various sources, although the guidance generally does not specically address the use of performance measures to determine the eectiveness of work zone technologies. A primer to help DOTs develop and monitor suitable work zone performance measures is available from FHWA (Ullman et al. 2011). e primer presents information regarding data needs and sources (Table 4) and reasons and considerations for using various performance measures for expo- sure, safety, and mobility. Guidance regarding safety performance measures for work zones, including processes for development, example performance measures, possible data sources, Source: Brown et al. 2018. Figure 32. AFAD with CMS evaluated in Missouri DOT study.

Literature Review 39   Source: Liao 2014. Figure 33. Flowchart showing process used by smartphone app for work zone navigation for the visually impaired (GUI = graphical user interface). Data Needs Sources of Data Traffic Crashes • Time, location, and direction of travel • Severity • Manner of collision • Contributing factors • Statewide crash database • Manually collected crash reports from the local police agency Worker Accidents and Injuries • Time, location, and activity • Type (i.e., injuries involving traffic) • Severity • Agency occupational safety division records • Agency-established injury reporting system • U.S. Department of Labor Bureau of Labor Statistics • National Institute of Occupational Safety and Health Fatality Assessment and Control Evaluations • Highway contractor injury records Agency Work Zone Inspection Scores • Rating scores • Agency work zone inspections or field reviews Service Patrol/Emergency Management Services (EMS) Dispatch to Project Location • Time and location of dispatches • Type of response (for service patrol dispatches) • Agency services patrol dispatch logs • EMS dispatch logs Table 4. Data needs and sources for work zone safety performance measures.

40 Use of Smart Work Zone Technologies for Improving Work Zone Safety and methods for collecting data, is available from ATSSA (n.d.). A publication from ATSSA and FHWA provides direction on collecting and analyzing work zone safety data (Chandler et al. 2013). A research study by Hallmark et al. (2013) synthesized existing DOT practices regarding the use of work zone performance measures for mobility and safety, including tables of perfor- mance measures used by different DOTs and example strategies for the communication of work zone performance to the public. Development of Performance Measures for Work Zones The Virginia DOT has undertaken research studies to facilitate the development of DOT- specific performance measures for work zones. A research study by Fontaine et al. (2014) investigated various facets of a possible program for work zone mobility performance mea- sures for the Virginia DOT, including determining potential performance measures and thresholds and developing recommendations for data sources. In another research study by Kweon et al. (2016), recommendations were developed for four performance measures and eight summary measures for use by the Virginia DOT in assessing work zone safety (Table 5). The study also found that the inclusion of exposure measures can affect safety assessments for work zones. Example Applications of Data Collection Technologies to Support Performance Measures for Work Zones In some instances, data collection technologies such as probe vehicles and big data analytics are used to support work zone performance measures. In a research study by Kamyab et al. (2019), probe vehicle data were used to develop a template for a Work Zone Mobility Audit, and 250 work zones in southeast Michigan were evaluated based on metrics developed for Measures of Work Zone Safety Type Recommended for Performance Measures 1 Total Crashes Count Measure 2 Fatal and Injury Crashes 3 Crashes per Work-Zone-Hour-Mile Rate Measure 4 Fatal and Injury Crashes per Work-Zone-Hour-Mile Recommended for Summary Measures 1 Fatal and Injury Crash Victims Count Measure 2 Work Zones Exposure 3 Work-Zone Hours 4 Work-Zone Miles 5 Work-Zone-Hour-Miles 6 Crashes per Work Zone Rate Measure7 Crashes per Work-Zone-Hour 8 Fatal and Injury Crashes per Work Zone Table 5. Work zone safety performance measures recommended for use in Virginia (copyright © 2016 the Commonwealth of Virginia).

Literature Review 41   individual work zones. A circular from FHWA (2019b) provides an overview of the Virginia DOT’s practices for using performance measures obtained from probe vehicles to determine work zone mobility. A report by Sadabadi et al. (2016) provides an overview of the Work Zone Performance Monitoring Application developed for the Maryland State Highway Admin- istration in response to FHWA requirements for work zone safety and mobility. The report describes example applications of the system (Figure 34) and discusses its integration with the Regional Integrated Transportation Information System (RITIS). Wang et al. (2017b) presented a case example of the Iowa DOT’s Traffic Critical Projects program for performance visualiza- tion of work zone projects based on radar sensor data. The program uses big-data methods to analyze the data and efficiently display it in an accessible web-based interface to help inform DOT decisions. Summary of Literature Review Findings and Resources for Smart Work Zone Technologies A summary of literature findings and resources for smart work zone technologies is shown in Table 6. A more extensive table that includes survey results and additional smart work zone technologies is provided in Appendix I. Source: Sadabadi et al. 2016. Figure 34. Graphic showing congestion and alerts from Maryland Work Zone Performance Monitoring Application on I-70 work zone.

42 Use of Smart Work Zone Technologies for Improving Work Zone Safety Smart Work Zone Technology Example Findings from Literature Example Resources General • Evaluation framework developed and used to determine benefit-cost ratios (2.1 to 1 to 6.9 to 1) (Edara et al. 2013a) • 68% of 147 work zone safety technology studies concern speed reduction systems (Nnaji et al. 2020) • FHWA Work Zone Intelligent Transportation Systems (ITS) Implementation Guide (Ullman et al. 2014) • FHWA Work Zone ITS Implementation Tool (GitHub 2021b) • Connecticut DOT Smart Work Zones Guide (IBI Group 2017) • Smart Work Zone Design Standards (Massachusetts DOT 2016c) • Smart Work Zone Standard Operating Procedures (Massachusetts DOT 2016a) Traveler Information Systems • Improved speed and lane deviation for smartphone messages compared to the PCMS (Craig et al. 2017) • Reduced vehicle speeds (1.25 mph to 3.64 mph for DMSs) (Edara et al. 2011) • Reduced vehicle speeds (13% to 17%) for graphic-aided PCMSs (Huang and Bai 2014) • Improved time-to-collision and deceleration for connected vehicle messages (Yang et al. 2020) • Oregon DOT Portable Changeable Message Sign Handbook (2nd Edition) (King and McCrea 2018) • Job special provision (15-32) (Missouri DOT 2018) • Guidelines on Improving Work Zone Safety Through Public Information and Traveler Information (Roadway Safety Consortium 2011) Queue Warning • Reductions in crashes (22%) and near crashes (54%) (Hourdos et al. 2017) • Crash reductions: 15% (total) and 63% (injury) (Schulze 2018) • CMFs for nighttime lane closures with portable rumble strips: 0.717 (non-queuing conditions) and 0.468 (queueing conditions) (Ullman et al. 2018a) • Special provision 12RC812-A705-02 (Michigan DOT 2021) • Concept of operations (Minnesota DOT 2015a) • System requirements (Minnesota DOT 2018) • Long-term typical application (Minnesota DOT 2021b) • Traffic Engineering Manual (Part 6, Section 640-29.1 and Plan Note 642-57) (Ohio DOT 2021) • Traffic standards [WZ-ITS(1)-19 and WZ-ITS(3)-19] (Texas DOT 2019) • Plan sheet library (TC161, TC162, TC165, and TC166) (Washington State DOT 2021) Work Zone Data Collection Technologies • Waze identified incidents 10 minutes earlier than traditional approaches (Amin-Naseri et al. 2018) • DOTs at different levels of development (Cheng et al. 2017) • One crash for every 147 hard-braking events (Desai et al. 2020) • Work Zone Safety Data Collection and Analysis Guide (Chandler et al. 2013) • Work Zone Data Exchange (WZDx) Specification (GitHub 2021a) • A Framework for Work Zone Activity Data Collection and Management (Version 3) (Stephens et al. 2019) Dynamic Lane Merge • Beneficial when value of travel time greater than $4.85/hour (Datta et al. 2007) • Three fewer crashes per week (Kansas DOT 2016a) • Maximum queue length reduced from 8 to 2 miles (North Carolina DOT 2019) • Guidance for the Use of Dynamic Lane Merging Strategies (ATSSA 2012) • Boilerplate special provision (Minnesota DOT 2020b) Dynamic (Variable) Speed Limit • Average speed reduction of 2.2 mph and speed compliance eight times higher for uncongested work zone (Edara et al. 2013b) • Maximum reduction in mean speed of 4.7 mph (Mekker et al. 2016) • Average speeds 15 to 25 mph lower than original speed limit (Van Jura et al. 2018) • Guidelines for Work Zone Variable Speed Limits (North Carolina DOT 2011) • Standard Construction Drawing (MT-104.10) (Ohio DOT 2018) Table 6. Summary of literature findings and resources for smart work zone technologies.

Literature Review 43   Smart Work Zone Technology Example Findings from Literature Example Resources Work Zone Location Technologies • Smart Work Zone Activity App (SWiZAPP) successfully tested for user-friendliness and functionality (Adu-Gyamfi et al. 2019) • Development of Statewide Work Zone Information System (SWIS) for Minnesota with need for further field testing (Parikh et al. 2019) • Electronic Reference Library (Sections 486.12, 2528, and 4188 for Smart Arrow Board) (Iowa DOT 2021a) • Concept of Operations and Requirements for Smart Arrow Board (Athey Creek Consultants 2018) Notification of Construction Equipment Entering/Exiting - • Use of Smart Work Zone Technology to Improve Work Space Access Point Safety (American Road & Transportation Builders Association 2019) • Boilerplate special provision (Minnesota DOT 2020b) • Long-term typical application (Minnesota DOT 2021b) • Standard Construction Drawing (MT-103.10) (Ohio DOT 2018) • Traffic Engineering Manual (Part 6, Section 640-29.2 and Plan Note 642-59) (Ohio DOT 2021) Work Zone Intrusion Alarm • Successful trials for two systems (Khan et al. 2019) • Successful field experimental trials for two systems and development of selection guide (Marks et al. 2017) • Mobile work zone alarms increased merge distances up to 122 ft (Brown et al. 2015) • System requirements (Minnesota DOT 2015b) • Long-term typical application (Minnesota DOT 2021b) Proximity Alert System • Accurate worker location (within 6 ft) and VMSs posted speed limits and messages based on worker presence (Banaeiyan et al. 2016) • Bluetooth better performance than RFID or magnetic field (Cho et al. 2017) • Signals understood by participants (Park and Sakhakarmi 2019) - Radar Speed Feedback Signs • 30% reduction in deceleration rates (Hourdos et al. 2019) • Lower vehicle speeds (0.7 to 5.6 mph) and speed variance (Jafarnejad et al. 2017) • 50% reduction in drivers exceeding speed limit by at least 15 mph (Roberts and Smaglik 2014) • Median speed reduced by 4 to 13 mph with presence lighting (Sakhare et al. 2021) • Special provision [12RC812 (A685)] (Michigan DOT 2021) • Long-term typical application (Minnesota DOT 2021b) • Construction Manual (Chapter 15) (South Dakota DOT 2020) Table 6. (Continued).

44 A survey was developed and administered in order to gain greater understanding of the state of the practice for the use of smart work zone technologies in the United States. The survey, which included 18 questions, was reviewed by the topic panel before being sent to each state DOT via Qualtrics Survey Software (Qualtrics 2021). The survey was sent to one recipient in each DOT. The contact list for the survey was developed based on information obtained from the FHWA, and an effort was made to identify the appropriate person at each DOT to complete the survey. In addition, respondents were encouraged to collaborate with others at their DOT and to forward the survey to the staff member who would be most capable of answering the questions and providing the most accurate information. Responses were received from all 50 state DOTs and the District DOT for a 100% response rate. Survey questions asked specifically about eight smart work zone technologies [traveler infor- mation systems, queue warning, dynamic lane merge, dynamic (variable) speed limit, work zone data collection technologies, work zone location technologies, work zone intrusion alarm, and notification of construction equipment entering or exiting]. Each question asking about these specific smart work zone technologies also included an “other” option with space for respondents to provide information about technologies not listed on the survey. Topics that were covered by the survey include frequency of use of smart work zone technologies and system components, performance, factors considered, implementation considerations, policies and guidance, perfor- mance measures, and evaluation studies. A copy of the full survey can be found in Appendix A. A list of responding DOTs is shown in Appendix B, and the survey responses by DOT, including comments and resources submitted, are shown in Appendix C. This chapter is organized into the following sections: DOT Use of Smart Work Zone Tech- nologies (Survey Questions 1, 2, 4), Performance of Smart Work Zone Technologies (Survey Questions 3, 12–16), Components for Smart Work Zone Technologies (Survey Questions 5–7, 9), DOT Implemen tation Considerations for Smart Work Zone Technologies (Survey Questions 8, 10, 11), and Other Survey Feedback from DOTs (Survey Questions 17, 18). DOT Use of Smart Work Zone Technologies DOT Implementation Status of Smart Work Zone Technologies Question 1 of the survey sought information the implementation status (already implemented, not currently implemented but plan to implement in future, or no current plans to implement) for various smart work zone technologies. As shown in Table 7, traveler information systems are the most commonly used smart work zone technology, and more than half of DOTs have implemented traveler information systems and QWSs. In addition, approximately half of the DOTs plan to use work zone location technologies in the future. The majority of DOTs do C H A P T E R 3 Survey Results

Survey Results 45   not currently plan to employ work zone intrusion alarms. Other types of smart work zone technologies that DOTs have implemented include a website to provide information about lane closures, dynamic ramp meter with portable traffic signal, automated speed enforcement, and a downstream speed-notification system. Some DOTs have future plans to implement other smart work zone technologies such as license plate readers and overheight vehicle warning systems. Maps showing the geographic distribution of answers by DOT for traveler information systems, queue warning, work zone data collection technologies, and work zone location technologies are shown in Figure 35 through Figure 38. Some general trends on the maps can be observed. The use of QWSs and work zone data collection technologies is more prevalent in the Midwestern, North Central, and South Central DOTs. The geographic locations of DOTs that have imple- mented traveler information systems or plan to implement work zone location technologies in the future are distributed throughout all regions of the United States. Some DOTs provided additional details in the comments regarding their implementation of smart work zone technologies. A full list of survey comments for each question may be found in Appendix C. Some of the notable comments are summarized in the following list (DOT names are not provided in order to preserve the confidentiality of survey comments). • Two DOTs indicated that some of the technologies that they have tried are not used regularly. • Three DOTs are piloting various smart work zone technologies. • One DOT is in the process of purchasing devices to provide real-time locations of work zones on Interstates. Q1. What is the current status of implementation of each of the following smart work zone technologies by your agency? Smart Work Zone Technology Already Implemented (%) Not Currently Implemented but Plan to Implement in Future (%) No Current Plans to Implement No Response Traveler Information Systems 78% 10% 12% 0% Queue Warning 61% 18% 20% 2% Work Zone Data Collection Technologies 33% 37% 27% 2% Dynamic Lane Merge 27% 31% 39% 2% Dynamic (Variable) Speed Limit 24% 27% 47% 2% Work Zone Location Technologies 24% 51% 24% 2% Notification of Construction Equipment Entering/Exiting 18% 33% 47% 2% Other 10% 6% 4% 80% Work Zone Intrusion Alarm 6% 33% 59% 2% NOTES: Sort order = already implemented (high to low); cell shading based on increments of 25%; total number of respondents = 51. Table 7. Survey results for implementation status of smart work zone technologies.