Data on traffic signal systems is usually best known at local or regional levels and is rarely shared across multiple states. In an attempt to better understand some of the basic characteristics of traffic signal systems in North America, I developed a simple overview survey of traffic signal maintenance and design elements (see attached survey form). As ITE District 6 Technical Chair, working with Nazir Lalani and the Technical Council of ITE, this survey was distributed across the United States and Canada. Near 150 responses were faxed back in the spring of 1996 and the data are summarized in the following tables. Responses covered 38 states in the US and six provinces of Canada. The responses are organized by ITE District 6 agencies, other US agencies and Canadian agencies. The following tables summarize some statistics from the survey, followed by general overall observations. (Detailed summaries of Western US and North America data is available).

Four key characteristic elements of signal systems from the survey (controller, detection, phasing and mounting) are dominated by NEMA controllers, loop detection, permissive phasing and mast arm mounting. The majority of agencies use NEMA controllers. Most agencies are either NEMA or 170 dominant. Few agencies mix their equipment. Many mid-west and eastern cities still have large percentages of electro-mechanical controllers remaining in service.

The vast majority of agencies use loop detection for their traffic signals and other forms of detection are generally past experiment or current demonstrations. The City of Troy Michigan has made significant investment in video detection. However, in general, the use of non-loop detection has little widespread application. Agencies use a wide variety of loop shapes. Typically, they are most common among regions (for example, Florida is dominated by rectangular loop users). The rectangular shape is the most frequent loop design from the survey (typically noted as quadrapoles). Circular loop design is a more recent trend which is primarily noted in Oregon and California (in Oregon it is a factor of the contractors having the equipment to install circular saw cuts which has saved them construction cost). Many agencies place detection loops into the crosswalk area (68 percent). This is most common among rectangular loop users. It is clear from the survey that agencies in western states use more loops in their designs than their eastern counterparts. The fact that the median loop design from the survey is a single loop per lane, but the average number of loops per lane is higher (about two per lane for left and through lanes) is a result of this higher number of loops in the west.

Permissive traffic signal phasing is used on nearly 50 percent of the signals in the survey. Compared to other aspects of the survey, traffic signal phasing is spread out among three primary types (permissive, protected and protected/permissive (P/P)). Additionally, some agencies use large percentages of protected/permissive phasing or protected phasing. Split phasing is the least used (only represents about six percent of signals.

By over a two-to-one ratio, agencies are designing more mast arm signals than span wire signals. Out of 149 agencies responding, 105 use mast arms. Many of the responses that indicated they had designed span wire traffic signals noted that they were temporary installation. In the west, virtually all the agencies design their permanent signals with mast arms.

Meeting the MUTCD desire of 12 inch signal indications, nearly 70 percent of the more than 30,000 traffic signals represented in this survey use 12 inch aspects. Use of programmed visibility heads is low (less than five percent) and the use of LED indications appears to be growing rapidly (nearly four percent for technology which is

relatively new). Seventeen of the agencies have over 10 percent of their signals with LED aspects and a few have taken a big plunge into LED’s (over 50 percent). Those interested in LED use should look to these agencies for performance data and results in term of maintenance and operating costs.

The vast majority agencies have routine maintenance programs for their traffic signals (nearly 90 percent). Most undertake maintenance using in-house forces (68% in-house, 18% contract maintenance and 14% combine in-house and contract maintenance). An average traffic signals receives about 5 to 6 scheduled field visits per year and about the same number of emergency calls per year. Relamping (if scheduled) is done once average every 14 months. The survey presents some very interesting approaches to maintenance that are worth reviewing. The common belief that routine signal maintenance "costs too much" (particularly to agencies who do not do it) is completely blown away by several agencies who undertake rigorous programs and at low cost per signal. The average operation and maintenance cost per signal was about $2,670. While some of the agencies may not have included all costs, it is clear there are many agencies who are doing it better for less than others. There is not a correlation to routine maintenance and higher cost. Some agencies with limited programs have high annual maintenance costs while others who routinely visit signals have lower costs. It is clear that the profession could learn from various agencies ability to do it "quicker, faster,.better". Considering the variations in O/M costs, while a $500 variation per signal does not appear great, but if applied to all the signals in the survey it would amount to $15 million a year. This would make a great traffic signal system funding program!

Detection maintenance is a significant issue with many agencies. It is clear from the survey that agencies would like to get twice the life out of their loops compared to what they get today. The average loop life was about seven years, but the desired life averaged over 13 years. In most cases, agencies want their loops not to be replaced until pavement reconstruction occurs (note that in survey responses, if ranges of data were provided, typically the average was recorded in the summary - with loop life "forever" was recoded as 20 years). Recent trends in loops designs (pre-formed loops and circular designs) appear to be meeting this need. For those who replace loops frequently, it is worth contacting agencies who do not. Yes, there are several agencies who are happy with the life of their loops. Additionally, stated loop life and satisfaction span a wide range of weather/temperature conditions. This may indicate that some of the problems with loop life may be in specification and installation practices of local areas.

There is a wide range of what agencies are paying for new traffic signals. Starting from $18,000 in Wilmington, NC and Genesse County, MI to $200,000 in Las Vegas, NV (where they build one mile of interconnect with each signal), there is plenty of variation in what it costs to build a traffic signal. The average cost of a traffic signal in the survey was $75,000. Issues such as the number of loops and use of mast arm/span wire seem to have the greatest affect on cost. Western states with mast arms and multiple loops have higher construction costs. One thing is clear, the affect of modifying standard specifications developed in regions/states does not seem to affect the bottom line - resulting in reduced construction or maintenance costs. Cases where standard specifications are not changes or changed considerably produces a wide range of cost results (again pointing to an area where we should be able to do it "quicker, faster, better").

Communication is a significant cost to many systems. Just over 50 percent of the signals surveyed have communication with a central system or field master. This points to an area where future maintenance and operation costs can be reduced by lessening the need for labor intensive support to signal systems.

Beyond the basic comparisons between communities, the summary provides transportation engineers a prototype of what other agencies are able to accomplish within their budgets to meet traffic signal operational needs. While many cities are very aggressive in creatively maintaining and designing their traffic signal systems, other cities are not (in fact many still have substantial percentages of electro-mechanical equipment). All cities face increasing budget pressures and the survey shows that many agencies are clearly very innovative in addressing the public’s needs within their budgets.

This survey provides a platform for transportation professionals to perform further research in traffic signal maintenance and design and look for way to improve our efficiency in serving the public. The survey data will be placed on an ITE District 6 Internet web page for everyone’s use (for research and to allow new data or corrections to occur on-line). Follow up surveys in this area may be used to further share information about performance of system elements (such as signal systems and how they can allow higher level of operation and maintenance at lower cost). Rational for use of communication, multiple loops, LED’s, new detection or other traffic signal elements can be surveyed to better provide agencies with background for considering major capital investments which can produce overall cost savings. I greatly appreciate everyone’s participation in this survey and look forward to your help in any follow up survey in the future.

Ransford McCourt, P.E.
DKS Associates