Connected and Autonomous Vehicle Technology: Determining the Impact on State DOT Maintenance Programs (2024)

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.

69   Traffic Signal Controllers for CAV Applications 9.1 Description of Traffic Signal Controllers It is expected that connected vehicles (CVs) will ultimately integrate with traffic signals and rely on traffic signal controllers that generate signal phase and timing (SPaT) messages, including green, yellow, red, and the amount of time until the next phase. Early applications for intersection components had only included roadside units (RSUs) and dedicated short-range communication (DSRC). As a result, these applications were the same as those described in Chapter 4. Autono- mous vehicles (AVs) also need to be able to interpret traffic signal phases, usually through image recognition. Smart intersections were one application that had been piloted for connected and autonomous vehicles (CAVs) and will require additional components. Smart intersections use an array of infrastructure-mounted sensors for object detection. The systems allow intersections to adapt and better manage demand. They can function independent of CAVs, but are also capable of broadcasting messaging. In addition to regular equipment, the signal cabinet at a smart inter- section includes a plurality of equipment depending on the intended applications. The equipment in the smart signal cabinet includes the following: • Automatic traffic controller (ATC): e.g., Siemens 2070 traffic signal controller or Cobalt rack mount controller. • ATC processor [or central processing unit (CPU)]: For example, 1C CPU was used for ATC in a Salem, Oregon signal phase and timing (SPaT) Challenge Project. This type of CPU is commonly used with a 2070 traffic signal controller. • Memory: This is used to store security certificates, application data, etc. • Multi-modal intelligent traffic signal system (MMITSS) and MMITSS roadside processor (MRP): e.g., the Savari StreetWave processor was used in Arizona (Multi-Modal Intelligent Traffic Signal System: System Design—California Portion 2016). • Managed field Ethernet switch (MFES) or Ethernet switch. • GPS receiver and antenna reference point (ARP). • RSU: DSRC or cellular. • Power supply: A centralized power supply assembly able to provide power to all units or appropriate power supplies for the ATC, ATC processor, and wireless communication device’s processing units. All controllers already have a source of power. • Input/output (I/O) assembly. • Power over Ethernet (PoE) switch and required accessories such as cables and splitters. • Message processor for roadside equipment (RSE). • Backhaul modem. • Wireless transmitter-receiver unit other than RSU: This unit is required for communication between the ATC and sensors, lights, audio devices, etc. These are short-range radio access channels with 20 to 300 m of range and work in frequencies at about 430 to 900 MHz. C H A P T E R 9

70 Connected and Autonomous Vehicle Technology: Determining the Impact on State DOT Maintenance Programs • Sensor controller or processing units: These vary depending on the type, configuration, and number of sensors. • Cameras or other sensors: Applications such as pedestrian detection require video, thermal, or other types of cameras. 9.2 Gathering Maintenance Information for Traffic Signals for CAVs Information was initially gathered through the survey of state department of transportation (DOT) maintenance practices, interviews with state DOTs and cities, contact with vendors, and interviews with maintenance contractors (as described in detail in Chapter 2). Information gained during Phase I was presented to the NCHRP panel at the end of Phase I. After discussions at the conclusion of Phase I, panel members showed interest in gathering any new information that was available about maintenance of traffic signals, but traffic signals were not the focus of subsequent interviews. As a result, an additional literature review was conducted, and all the available information is summarized in the sections that follow. 9.3 Examples of Applications of Traffic Signals for CAVs Of the 39 DOTs that responded to the agency survey, five indicated they had added, six had increased, and six had modified traffic signal controllers to accommodate CAVs. In the next 3 years, six planned to add, eight planned to increase, and 12 planned to modify traffic signal controllers to accommodate CAVs. In total, 10 DOTs were questioned about a subset of assets that they had implemented to address CAVs, and three were asked about implementation of traffic signal controllers (Florida, Kentucky, and Minnesota). Responses are summarized in the subsections that follow. 9.3.1 Florida Traffic Signals for CAVs Florida DOT (FDOT) was spending roughly $50 million per year on maintenance of 8,600 traffic signals and was adding CAV equipment as the signal equipment was maintained. The state had an arrangement where the DOT provided the maintenance funding, but individual municipalities operated the traffic signals. 9.3.2 Kentucky Traffic Signals for CAVs The Kentucky Transportation Cabinet (KYTC) had switched to automatic traffic controllers (ATCs) (Intelight 2070 ATCs running MaxTime). The controllers were being managed through a central software suite called MaxView. The decision had been made to replace aging 170 Wapiti controller architecture with an upgraded model. The controllers were being implemented statewide. The decision to replace the aging 170 Wapiti controller architecture was not to accommodate CAVs, but to help facilitate integration of CAV technology. The decision for placing controllers was made primarily by their central office working with the districts to select the corridors. The KYTC had been implementing 2,070 controllers over 5 years. The new controllers were Intelight 2070 ATCs running MaxTime. The DSRC radios were WAVEMOBILE Fiberwire 8011 RSU/OBU Radio.

Traffic Signal Controllers for CAV Applications 71   9.3.3 Minnesota Traffic Signals for CAVs Minnesota DOT (MnDOT)’s first deployment of CAVs was on the TH 55 corridor for the American Association of State Highway and Transportation Officials (AASHTO) SPaT Challenge. The deployment included 22 intersections from downtown Minneapolis to the west to the I-494 corridor. The earliest discussions had occurred in 2015 and considered manufacturer solutions that were viewed as a black box. The goal was to learn about DSRC. At this stage, they had 18 months of experience transmitting SPaT and MAP messages. They were connecting this message report- ing to data portals (Greenhill was the vendor). The TH 55 corridor deployment used an Intelight controller, although the project started with Econolite. Intelight MaxView was the MnDOT traffic or transportation management center (TMC) central software. The agency had been looking to improve the central software to provide application programming interface (API) access or feed a data portal with SPaT/MAP message data. Another deployment was planned with newer technology for Smart Snelling Avenue— a deployment of 13 units with cellular communication that could be read or accessed using a smartphone. The Smart Snelling deployment would use Econolite. Across both corridors, the state was using two standard vendors for traffic signals statewide. From experience with these two vendors, MnDOT was seeing CAV-ready controllers becoming standard. Both corridors were looking at a priority solution for mobility. MnDOT was focusing their priority on snowplows with on-board units (OBUs) for signal request messaging. The snowplows only had test priority at a demo of four intersections. However, MnDOT had 22 snowplows equipped with DSRC from their research budget. The vehicles could perform vehicle-to-vehicle (V2V) communications and had been focused on lane-keeping solutions, with research currently happening on a MnDOT-owned or sponsored test track. The snowplows were also part of an agency weather information decision support system that was pulling data back to the central office, fusing and processing data and decision support, and then recommending treatments to the fleet. Financial backing came from state highway funding, and more specifically, a split between highway funding and research dollars institutionalized over time based on standard intelligent transportation system (ITS). The TH 55 corridor improvements cost $1 million and the Smart Snelling deployment cost $0.5 million. 9.4 Maintenance Needs for Traffic Signals for CAVs Short- and mid-term maintenance needs will be subject to the adoption of CAV components by agency operators. Electronics have shorter wear cycles compared to rugged transportation infrastructure. A higher level of maintenance attention may be required to address equipment failures and upgrades to ensure they are adequate for CAV needs. Maintenance standards were unknown at the time. The surveyed states that responded about the use of traffic signal controllers were questioned about maintenance needs or practices. Their responses are summarized in the following subsections. 9.4.1 Kentucky Traffic Signal Maintenance Needs for CAVs KYTC controllers were being replaced as needed at a material cost of about $2,500 per controller. The labor involved was probably about 1 to 2 person-hours per controller replacement. No performance measures were noted, and no changes had been made to data collection practices.

72 Connected and Autonomous Vehicle Technology: Determining the Impact on State DOT Maintenance Programs 9.4.2 Michigan Traffic Signal Maintenance Needs for CAVs Michigan DOT (MDOT) had increased communications capacity for RSUs. Mainly, they had added switches and cell modems to existing sites to power RSUs. They noted that fiber was not available in several areas, so if a location was not already connected, they were adding modems. Signal connectivity allowed them to obtain data, which could be analyzed remotely. Use of the connectivity also allowed MDOT to conduct traffic signal maintenance and fix other issues (e.g., drift) remotely. Ultimately, they planned to change signal timing “on the fly.” They had recently contracted with a vendor to assist them with the process. The long-term plan was to have 3,200 signals with communications capability, with 500 going live in 2021. RSUs had been deployed and were available for vehicles that were able to connect. However, use of the RSUs seemed to be limited by the available applications (e.g., weather specific). Additionally, their snowplows with connectivity options were not using the RSUs. They were doing creative TIM messaging and relaying that information to the associated dynamic message sign (DMS). 9.4.3 Minnesota Traffic Signal Maintenance Needs for CAVs Signals on the TH 55 corridor had limited maintenance needs to date. MnDOT worked with support from the University of Minnesota. A consultant also provided support for planning and technical support of the installation. The agency direction had been to keep as much work in- house as possible. Both the metro signal and maintenance staff had worked on this equipment. Maintenance could include the replacement of cabinets and restoration of electrical service from knockdowns. RSU antennas were normally installed on 4 ft mounting poles above the mast arms. MnDOT had documented a location mounted lower to achieve line of sight if it was hit by a vehicle. MnDOT also had health monitoring software which determined whether locations had communication and power. The health check was being performed once a day. It had resulted in some power resets and the need to push out updates (with three updates needed in 18 months). Contractors had provided support to agency forces that would have long-term, hands-on responsibility. 9.5 Standards, Guidelines, or Best Practices for Application and Maintenance of Traffic Signals for CAVs No standards or guidelines for traffic signal maintenance were found specific to CAV needs. A number of recommendations, which were relevant to both maintenance and placement, had been made. Hietpas and White (2019) assessed CAV infrastructure needs and indicated the following needs for traffic signal infrastructure: • Install traffic signal controllers with SPaT capabilities, • Assess whether larger traffic cabinets are needed for the future, • Address electrical code issues for conduit for electrical conductors and Ethernet cabling separately, and • Develop standards to assess small cell installation or other technology. One of the most common recommendations relevant to traffic signals was consistency. Recog- nition of traffic signal presence and state was more complicated than for pavement markings and signing, but less information is available. The National Committee on Uniform Traffic Control Devices (NCUTCD) conducted a survey of the automotive industry to support CAV deployment 16818-02_Ch06-13-3rdPgs.indd 72 2/26/24 11:24 AM

Traffic Signal Controllers for CAV Applications 73   and made the following recommendations for traffic signals (NCUTCD 2019; Chan and Wang 2021; Carlson 2021): • Signals should be uniformly placed; horizontal traffic signals are particularly problematic as shown in Figure 9-1. • Signals should be standardized, including position, location, color, shape, and refresh rate. • Back plates may be beneficial for east or west placement, particularly in low sun conditions. • Signals should have a clear, unambiguous association with a specific lane. • High and low brightness should be standardized. • A 12 in. diameter signal head is preferred over an 8 in. signal head. • Signals that target different classes of vehicles (e.g., cyclist or bus signals) should be placed and located at sufficient distance from each other so that their individual applications can be differentiated. • Use green lights rather than flashing beacons where possible (e.g., pedestrian crossing control is better as standard green-yellow-red lights rather than flashing red), and STOP and GO directives should be explicit. 9.6 Workforce Needs for Maintenance of Traffic Signals for CAVs Traffic signal workforce skills already exist in some agencies, but higher adoption of smart intersections could drive up the demands on maintenance staff with these skills. Additional workforce and appropriate training may be a required strategy for agencies leading the adoption of CAV technology and smart intersections. Additionally, many agencies may contract out maintenance for some electronics-based assets. 9.7 Summary of Maintenance Needs for Traffic Signals for CAVs It is expected that CVs will ultimately integrate with traffic signals and rely on traffic signal controllers that generate SPaT messages. Early applications for intersection components have only included RSU and communications needs. As a result, maintenance needs are primarily specific to those assets. However, some recommendations had been made specific to AVs, which need to interpret traffic signal state. Recommendations to accommodate AVs were similar to those for signing and included consistency in placement and maintenance. Source: Mike Kuhlman/Shutterstock.com (left); viphotos/Shutterstock.com (right). Figure 9-1. Example of inconsistent signal placement.