Office of Operations
21st Century Operations Using 21st Century Technologies

A Methodology and Case Study: Evaluating the Benefits and Costs of Implementing Automated Traffic Signal Performance

Chapter 4. Conclusions and Recommendations

This chapter presents summary findings of the six case studies. More details about these case studies can be found in Appendix A: Case Studies. These case studies summarize agencies' experiences from the three major points of their automated traffic signal performance measures (ATSPM) implementations: planning for the system, procuring and deploying the system, and operating the system. Table 8 shows an overview of important operational characteristics of the agencies.

Table 8. Major agencies' characteristics.
Agency Number of Signals Number of Signal Operations Staff Use of Automated Traffic Signal Performance Measures Type of Deployment
Clark County
  • 98 traffic signals.
  • 3 HAWK signals.
  • 24 signals for other agencies.
  • 1 engineering manager.
  • 1 signal engineer.
  • 1 intelligent transportation systems engineer
  • 6 signal technicians.
  • Consultants (as needed — project specific).
As needed (now); continual (future). Trafficware
Cranberry Township
  • 49 signals.
  • 4 engineers.
  • 1 technician.
In early deployment. Econolite + Edaptive
Georgia Department of Transportation
  • 6,804 (all capable of collecting high-resolution data, approximately 80% configured to create reports).
  • 70–80 full-time employees, including consultants.
Continuous. UDOT open-source ATSPM software.
Lake County Department of Transportation
  • 180 signals; 133 signals under ATSPM, the rest will be under ATSPM by the end of 2019; 100% of signals have stop-bar detection.
  • 100% of signals have advanced mainline detection.
  • 87% of signals have advanced detection on all approaches.
  • 1.5 full-time employees, excluding sporadic engagement consultants.
On a daily basis. UDOT open-source ATSPM software version 4.0.2 (looking to replace with vendor solution).
Maricopa County Department of Transportation
  • 170 (117 monitored with ATSPM) operated by MCDOT, over 3000 total in Maricopa County.
  • 3 on-call consultants for traffic operations.
  • 1 on-call systems integrator.
  • 1 full-time engineer.
  • 1 full-time analyst.
Automated alerts; as needed, mainly for responding to public calls. UDOT open-source software.
Pennsylvania Department of Transportation
  • 2
  • Many (not relevant).
To support others. Intelgiht MAXVIEW
Utah Department of Transportation
  • 2,120 traffic signals in Utah.
  • 1,252 traffic signals owned and operated by UDOT.
  • 63 HAWK signals in Utah.
  • 19 HAWK signals owned and operated by UDOT.
  • 1 engineering manager (management in both timing and maintenance)
  • 2 statewide timing engineers (timing only)
  • 4 region signal engineers (management, plan review; & a little maintenance & operations)
  • 4 statewide timing technicians (timing only)
  • 8 signal maintenance technicians (maintenance only)
  • 3 signal timing consultant firms (timing only)
Continuous. Utah DOT open-source ATSPM software.
ATSPM = automated traffic signal performance measures. MCDOT = Maricopa County Department of Transportation. UDOT = Utah Department of Transportation.

Planning for the System

Most agencies did not have a formal planning process for the ATSPM system. They gained information about the technology through outreach from the Federal Highway Administration (FHWA) and the American Association of State Highway and Transportation Officials (AASHTO) Innovation Initiative or pooled fund study and decided to deploy it. It is also worthwhile to note that FHWA promotes a systematic approach to implementation that considers the impact on the workforce, business processes, and appropriate use of systems engineering to inform design and procurement. Many downloaded the UDOT source code (and had internal staff install, configure, and maintain). The size of implementations varied from 50 to 5,000 signals. Once a system is deployed, adding intersections is pretty simple (given the appropriate infrastructure). For larger agencies, the approach was to equip and configure every intersection, which made the change in business processes easier. For smaller agencies, the approach was to start small and make sure they saw the benefit in using the system, then expand to different corridors or metrics as the need arose.

UDOT and GDOT are probably the only agencies, to date, that have taken a systematic approach to integrating ATSPM in their signal operations and maintenance efforts. This is not a surprise considering they have been leaders in ATSPM development and implementation. UDOT has committed significant resources to develop the software and continues to enhance and update the system. Its support for innovative methods like ATSPM has been crucial in developing the ATSPM. UDOT's executive staff vision to establish a world-class traffic signal system was further championed by agency staff, who saw a means to reach this goal by building the capability to monitor signal conditions with performance measures. This was a great symbiosis of visionary leadership and knowledgeable staff to recognize which technologies should be acquired, and also how to enhance and further develop such technologies for the benefit of real-world applications.

GDOT also had strong executive support for improving agency operations, which was capitalized through a formal program of improvements of the State's signal infrastructure, such as controllers, communications, and detection. GDOT included data-logging capability, based on the Indiana Department of Transportation (INDOT) ATSPM data enumerations, as a requirement of its controller specification. Both UDOT and GDOT represent examples where great leadership skills were aided by technologically capable staff in a resource-rich environment. Under such conditions their ATSPM applications have flourished, and they were able to make contributions that have not only benefited themselves, but also the broader traffic signal agency community.

On the other hand, a number of smaller traffic signal agencies did not have all the resources of their larger State counterparts. Such agencies have been planning for their signal operations and maintenance outside of the ATSPM scope, but they saw an opportunity and adjusted their plans to fit the framework of opportunities offered by ATSPM. Such examples are Clark County, WA; Lake County Department of Transportation (LCDOT), and MCDOT. All of them had previously developed plans for signal operations improvement before they learned about ATSPM, and they have been adjusting those plans to include ATSPM ever since the technology caught their attention. Clark County, for example, has been upgrading its signal system for the past 10 years, after a period of diminished attention. It has been making significant investments to improve operations of signalized intersections (e.g., controllers, detection, and communications) and felt that ATSPM is the right technology to apply on top of other efforts. The story of LCDOT is similar—a relatively small agency, with limited resources and very knowledgeable and resourceful staff, that has recognized ATSPM technology as the next building block. This is on top of already suitable communication and detection infrastructure, to reach the next level of signal monitoring and operations. What has been especially helpful in LCDOT's case is a smooth communication between upper management and signal operations, which removes many institutional barriers to implement ATSPM and similar technologies.

Similar to other local agencies, MCDOT also benefited from a previous history of agency investments in traffic control infrastructure. Particularly interesting for MCDOT's case is its regional partnership with AZTech, a regional traffic management agency in the Phoenix metropolitan area that guides application of intelligent transportation system (ITS) technologies for managing regional traffic. This partnership carefully integrates individual traffic management strategies and technologies for the region's benefit while preserving operational control protocols important to individual jurisdictions.

The combined case of PennDOT and Cranberry Township involves a somewhat different approach. While Cranberry Township's story is similar to the aforementioned counties (e.g., planning for improvement of signal operations but only recently adding ATSPM in the overall framework), the PennDOT story is slightly different. With practically no signals owned and a desire to help many small counties, townships, and boroughs manage their signals, PennDOT has dedicated a lot of effort to plan for better signal operations and especially improve coordination among various signal stakeholders. As a part of its Corridor Modernization program, PennDOT started to investigate the possibility of changing ownership of traffic signals on certain critical corridors. Thus, the ATSPM initiative has been seen (and supported) by PennDOT as another tool to help the State streamline signal improvement efforts and improve coordination with local stakeholders.

Of the seven agencies studied, only two are quantitatively tracking (not comprehensively) benefits and costs associated with ATSPM. Frequently, the cost appears to be absorbed within existing Corridor Improvement Program (CIP) projects (related to general traffic signal costs) or as part of typical staff duties. Benefits could be calculated if an agency specifically tracked the number of hours and costs associated with operation and maintenance in the before-and-after ATSPM condition. Driver benefits (reduced delay due to a fixed detector or updated coordinated timings) are more difficult to quantify and may need a more detailed methodology for agencies to document or estimate.

Procuring and Deploying the System

It seems that UDOT's ATSPM open-source code is a good choice for larger agencies with adequate information technology (IT) resources, but may not be as effective for smaller agencies. Issues some agencies face are various—from inadequate resources to cope with constant hardware and software updates to having the latest version of source code to the need to customize source code, which often involves additional resources. This indicates that some of the interviewed agencies had issues with stability of the open-source software, but such stability is not seen as an issue at larger agencies (that have applied ATSPM software on a larger number of signals). The consequence of such issues is that many small agencies may find software as a service maintained by commercial vendors a better fit for their individual needs for signal monitoring based on the high-resolution performance measures. Some larger agencies like UDOT and GDOT have opened up their ATSPM systems to incorporate signals operated by local agencies within their States.

The systematic approach to transportation system management and operation employed by UDOT and GDOT allows the agencies to position resources and leverage strong integration between information technology and traffic management, and disseminate advancements to the broader community. UDOT's approach is particularly interesting because at the outset of its investment in developing the open-source software, it did not fully understand what benefits it would be able to achieve by using signal performance measures, yet it still risked committing those resources. Its ongoing efforts to maintain and update communication and detection systems and controllers enabled it to implement ATSPMs with relative ease. In many cases, data collection was simply enabled as a controller feature, while minimal changes to detector configurations were made in some cases. However, it is important to note that not every traffic signal in the State of Utah could immediately be integrated into the system, and even now, not every signal can report every metric. However, the base infrastructure now exists to enable those capabilities in future.

At the time GDOT initiated ATSPM deployment, it was in a good state with making investments in communication, detection, and updated controllers as a matter of maintaining its signal infrastructure. As mentioned earlier, GDOT made the high-resolution data-logging capability a requirement for procurement of traffic controllers. Implementation of ATSPM built upon the existing system with minimal investment. Like UDOT, GDOT was also able to fully commit to operations and maintenance of its own software and server, which resulted in quick familiarity with the system and contributions to the broader community. For example, GDOT invested resources in the first software and use-case ATSPM manuals, which were prepared by consultants to the general public's benefit. Similarly, soon after the first couple of versions of the ATSPM software were released, GDOT developed an add-on module that used some of the ATSPM data to calculate business-case ATSPM summaries, compiled under a tool called Measurement, Accuracy, and Reliability Kit (MARK 1). Although some parts of Georgia are well covered with commercial versions of ATSPM tools, GDOT actively participates in maintenance and enhancement of ATSPM open-source software.

Other smaller entities were not always able to stay with the open-source version for a long period of time. Clark County, for example, was getting ready for an ATSPM-like approach, which was evident because of its innovative use of logs and reports available in Trafficware, the county's central signal system used to manage its signal system. In doing so, it realized the importance of good detection and communications and continued to invest in signal infrastructure. It also installed a Bluetooth sensor system to collect and track travel time along different corridors. In 2013, it began working with Trafficware to develop a software application. This application would use high-resolution data to create reports, similar to what UDOT has developed. In 2014, Clark County asked Trafficware to create five reports so it could monitor operations based on their operational objectives. Clark County is one of those agencies that understood the limitations of owning and maintaining an open-source software, and, thus, went with a commercial version of ATSPM from the very beginning.

LCDOT, on the other hand, had a quite different approach. In short, it went through multiple ad hoc decisions and actions to test various options. It first installed a version of open-source ATSPM software on its own, and maintained it for a while. When a new version arrived, it understood how cumbersome installation could be and hired a local consultant to help with installation. However, even after successful installation it had issues with software crashes and limitations of adding and modifying features to address its specific needs. Finally, LCDOT decided to develop a request for proposal (RFP) and select a commercial solution from a vendor. In the RFP process, LCDOT staff took a proactive role and challenged potential vendors to propose and implement new functionalities that would address specific objectives of LCDOT's signal operations. The process went smoothly, and LCDOT has selected a vendor to supply a commercial ATSPM system, which has embedded a number of enhancements above and beyond the original UDOT source code.

MCDOT's path was perhaps more similar to the two DOTs than to the other smaller agencies. This is not surprising because Maricopa County itself is a larger entity than some States. Over the years, MCDOT has made a commitment to keep its traffic control infrastructure up to date and functional. For the ATSPM project, specifically, detection needs were largely met through a parallel effort to make signals ready for deployment of other advanced technologies. MCDOT is still in the initial phases of using the open-source ATSPM software, and so far it has not had server crashes or other strong reasons to consider commercial ATSPM implementations.

While both PennDOT and Cranberry Township could be considered early adopters of ATSPM, their approaches to ATSPM implementation differ from each other's quite significantly. While PennDOT's efforts are purely driven by a strategic desire to support various signal jurisdictions in the State (the State does not directly control any signals), it seems its early efforts with UDOT's open-source code have not led to a full success. After replacing the ATSPM platform (supported by PennDOT) with a commercial system (Intelight MAXVIEW), operations have stabilized, yet are still imperfect, and it is still uncertain how many local agencies in Pennsylvania have capitalized on this opportunity.

On the other hand, Cranberry Township has taken its own path, with some early tests through Purdue University followed by a fully commercial Econolite platform. In terms of system procurements, Cranberry Township even went one step ahead of many others by deploying a commercial adaptive traffic control (ATC) system based on ATSPM (Econolite Edaptive). However, neither the ATSPM nor Edaptive has been fully utilized—they are procured and installed but still not fully operational with clearly tracked benefits.

Operating the System

Most of the surveyed agencies use their ATSPM systems on a daily basis but their operational practices may not always be well integrated into their business models, which connect agencies' goals with operational objectives and field strategies. For example, if an agency has a clear plan for how to operate signals along a corridor, it is more likely the agency will be more successful documenting its operation success by using the ATSPM. On the other hand, if an agency does not know exactly how to use reported performance measures, an abundance of data from the ATSPM system will not necessarily have a significant impact on its day-to-day operations. On the other hand, some agencies do not have a "know-how" problem, but lack sufficient time to extract relevant information from the enormous amount of data ATSPM systems can provide, connect that information with potential implementable strategies, execute those strategies, and validate whether the activities had impact. While ATSPMs provide a useful tool to investigate problems already known, identifying issues with the most pressing need can be a bit like finding a needle in a haystack. Methods to improve or automate this process are the subject of current research.

As in many other aspects of ATSPM, UDOT has been a leader in actively using this technology. In this case, the agency's traffic signal staff immediately saw the utility of the metrics and identified ways to use them to solve many different problems in traffic signal operations and maintenance. Examples included new methods of evaluating signal timing, identifying maintenance issues, and triaging public complaint calls. To align support of its consultants with new ATSPM practices, UDOT has restructured its contracts to refocus consultant activities from the traditional process of signal retiming toward using ATSPMs. This effort has been successful, with many consultants in Utah now actively using ATSPMs.

Other agencies have sought to achieve similar successes through various means. Clark County, for example, wants to address the matter of identifying issues by setting a triggering mechanism that would detect abnormal operations (in a particular context) and report an issue. It plans to fine-tune such a mechanism once ATSPMs are enabled at all its intersections. Clark County developed a measure of effectiveness (MOE) framework for each corridor to determine the most important metrics given the context by creating "corridor atlases" to document operational issues and objectives at each intersection.

LCDOT still has not made full use of its ATSPM capabilities, but it quickly realized its resources will be better used if ATSPM operations are done by a commercial vendor rather than in-house. Thus, it decided to hire a consultant who will be responsible for both maintenance of the hardware, software, and routine daily operations of the ATSPM. In its previous practice it has confirmed (although not fully documented) proactive use of ATSPM reduces complaint calls and costs associated with infrastructure malfunctioning. LCDOT's in-house staff will still be responsible for decision making, but retrieval of relevant information from the abundance of ATSPM data will be done by a commercial vendor. This labor division is supposed to ensure efficient use of ATSPM over a long period of time.

MCDOT is an example of an agency that has seen successes in its early use of ATSPM. It identified three directions of transitioning its practices toward active management and operations: (1) the low level at which most of the metrics are reported (individual movements), (2) a need to orient performance measures toward specific tasks, and (3) a need for additional training to bring traffic management center (TMC) operators and other signal operations staff up to speed on using the metrics. To overcome these issues, MCDOT is waiting for further development of the open-source software by innovators to address the first two points (although it may explore solutions using its own consultants), and anticipates documentation and training activities for the third point.

In Pennsylvania, the two interviewed agencies have quite different paths for ATSPM operations. PennDOT, which does not interfere in daily operations of the signals, expects investment in a statewide commercial Intelight license will attract more local jurisdictions to use ATSPM. On the other hand, Cranberry Township is only a few months away from achieving some concrete benefits of its ATSPM and Centracs® Edaptive deployments. Cranberry Township may be the first system in the country where benefits of an adaptive system will be conveniently and easily measured through ATSPM, which may be a good example of how to merge the strengths of ATSPM's powerful graphics and reporting with automation of an adaptive traffic control system.