Table of Contents

Chapter 2 – Configuration Management and Transportation Management Systems – Current Practices

Introduction

Given the important role of CM in a complex system, it is important to understand how it is currently being used for transportation management systems. A survey was conducted in the spring of 2000 to gauge the use of CM by transportation agencies in the United States. Of the 38 responses the most striking result was that 62 percent of freeway management systems use CM, and only 27 percent of signal systems use CM. These results indicate a need to educate the TMS community about CM in order to realize a significant commitment to this valuable resource-saving activity. The survey also revealed that many of the complex TMSs in this country are not using a formal change control process. As stated in chapter 1, the lack of formal change control processes calls into question the very integrity of many of these systems.

This chapter details the results of the survey. The results are presented in terms of the primary survey sections: TMS characteristics, CM plan, CM process, CM organizational issues, benefits/costs of CM, and testimonials. Please note that this survey was originally conducted for an NCHRP Synthesis project. The full results of this project are published in NCHRP Synthesis 294 (2001).

Survey Results

TMS Characteristics

The first section of the survey intended to gauge the size and extent of the responding agency's TMS(s). Such information provided the background needed to identify trends in the use of CM.

The first question of the survey asked about the core functions provided by the TMS. Respondents were to check all that applied, and if a particular agency performed more than one function, then the sample size would increase accordingly. Counting each function independently increased the sample size from 38 to 42. The resulting functional descriptions of the systems were as follows: 20 freeway management systems (FMS), 15 traffic signal control systems (TSCS), 2 automatic toll collection systems (ATCS), and 5 tunnel control systems (TCS). Figure 2.1 illustrates the percentage share of the functional classes of systems.

Figure 2.1 – Functional Classes of Systems
Figure 2.1 D

Respondents also were asked to provide information concerning the size of their system(s). This information was provided in terms of the number of signalized intersections, the lane miles of coverage, the number of CCTV cameras, and the number of variable message signs depending on the functional class of the system. They also were asked about the number and type of detectors, the number of ramp meters, the number of lane control signals, the number of road weather sensors, and the number of toll tag readers. The purpose of these questions was not to collect large quantities of data describing system size, but rather to provide a "check" to ensure the responses regarding CM were not skewed towards one particular size or type of system. A review of the responses indicated that the survey achieved a representative sample of the range of system classes and sizes throughout the country.

A key finding of the survey was that a relatively low percentage of TMSs use CM. What was particularly notable is that only 27 percent of signal control systems reported using CM. Table 2.1 illustrates the use of CM according to the function of the TMS. For the purposes of this table, automatic toll collection systems and tunnel control systems were grouped with freeway management systems.

Table 2.1 – Use of CM by System Classification
empty cell Traffic Signal Control Systems Freeway Management Systems
Percentage of Systems Using CM 27% 62%

Another clear trend in the survey responses is that the likelihood of a TMS using CM is dependent on the size of the system. Larger systems are more likely to utilize CM, as seen in figure 2.2.

Figure 2.2 – CM Use by System Size
Figure 2.2 D

Transportation agencies have used different types of core system software in TMSs. The software can be classified as custom developed software, for which the agency either does or does not own the source code, or the software can be classified as off-the-shelf and is typically purchased from a vendor. Fourteen of the departments use custom software for which the agency owns the source code, and ten of the departments use custom software for which the agency does not own the source code. Eleven of the agencies use an off-the-shelf software package purchased from a vendor. Figure 2.3 illustrates the distribution of software types.

Figure 2.3 – Core System Software
Figure 2.3 D

Most of the core TMS software was purchased in the 1990s. To quantify the size of the software systems, the average size of the survey responses is 264,000 lines of code and 94 megabytes of executable code. Finally, it is interesting to note in figure 2.4 that agencies are much more likely to use CM if their system uses custom software. This likely reflects the fact that changes to a custom system are more likely (and feasible) in a custom system built specifically for the agency's requirements. Furthermore, this also reflects the wider use of CM in freeway management systems. Of the 11 off-the-shelf systems identified in figure 2.3, 8 are signal control systems, which is consistent with the finding that only 27 percent of signal control systems use CM.

Figure 2.4 – Use of CM Based on Software Class
Figure 2.4 D

CM Plan

Given the size and cost of the TMSs being operated in this country, many transportation agencies have realized the need to institute formal CM to control changes in the systems. However, many transportation agencies have not become aware of this need until after the development phase of a system, when system operation and maintenance begins. This is particularly evident in the general lack of formal CM plans in TMSs. As seen in figure 2.5, of the 15 agencies that use CM, only 4 have a formal CM plan.

Figure 2.5 – Percent of Agencies Using CM that Have a Formal CM Plan
Figure 2.5 D

The four agencies that do possess a formal plan report that the plan is very important in ensuring an effective CM process. On a scale of 0 to 10 (10 being highly effective and 0 being completely ineffective), the importance of the plan on the effectiveness of the CM process received an average rating of 7.25. Yet not one of these agencies required system contractors to deliver a CM plan in its request for proposals for the initial system, which reveals a potential disconnect between the development and operations/maintenance phases of TMSs.

CM plans can address several different areas, including, but not limited to CM organization, CM responsibilities, CM training, configuration identification, change control, configuration status accounting, and configuration auditing. The regions and their plans' respective elements are illustrated in table 2.2. Note that all plans address organization and configuration control, but only half of the plans address training, accounting, or auditing.

Table 2.2 – Elements of CM Plans
Region CM Organization CM Responsibilities CM Training Configuration Identification Change Control Status Accounting Configuration Auditing
Miami Checked Checked empty cell Checked Checked empty cell empty cell
Los Angeles Checked Checked Checked Checked Checked Checked Checked
Charlotte Checked empty cell Checked Checked Checked Checked empty cell
Georgia Checked Checked empty cell empty cell Checked empty cell Checked

The agencies used many resources in the development of their respective CM plans. These resources included, but were not limited to: IEEE Standards, Software Engineering Institute, DOD Standards, and sample CM plans. Table 2.3 shows the resources used by each region.

Table 2.3 – CM Plan Resources
Region IEEE Standards Software Engineering Institute DOD Standards Sample CM Plans
Miami empty cell Checked Checked empty cell
Los Angeles Checked empty cell empty cell Checked
Charlotte empty cell empty cell empty cell empty cell
Georgia Checked Checked empty cell Checked

For more information on configuration management planning, see chapter 4.

Both agency staff and consultants contributed to the creation of the CM plans. To provide some context in terms of the resources required to develop the plans, Miami reported spending 800 hours on its plan, while Georgia Department of Transportation (GDOT) spent over twice that amount of time. (The Georgia system, however, covers not only the City of Atlanta, but also 5 surrounding counties and the Metropolitan Atlanta Rapid Transit Authority (MARTA) for a coverage of more than 220 freeway miles). In terms of funds, Los Angeles invested $80,000 in its plan, while Georgia invested $193,000.

CM Process

The next section of the survey dealt with issues related to the CM process used by the transportation agencies. One of the first issues addressed was the type of tools used by agencies to support their CM processes. Figure 2.6 illustrates the tools used by transportation agencies and their relative frequency of use. It should be noted that the three respondents who reported Excel as their CM tool have used this spreadsheet package as a simple means to document configuration items. It does not provide the full functionality that the other tools include.

Figure 2.6 – CM Tools Used by Agencies
Figure 2.6 D

For more information on configuration management tools, see chapter 8.

The survey results reveal that at different stages of the TMSs life cycle, different organizations (the transportation agency, the agency's consultant, or the agency's contractor) led the CM process, as seen in figures 2.7, 2.8, 2.9, and 2.10. While the transportation agency is most likely to lead the CM process during the planning, operations, and maintenance phases, a consultant is usually responsible for CM during the design and development phases. Furthermore, some agencies preferred to use the system's contractor during the design and development phases. These facts illustrate a key challenge in TMS CM – coordinating multiple parties' involvement in CM throughout the life of the system.

Figure 2.7 – Lead Organization During Planning Phase
Figure 2.7 D
Figure 2.8 – Lead Organization During Design Phase
Figure 2.8 D
Figure 2.9 – Lead Organization During Development Phase
Figure 2.9 D
Figure 2.10 – Lead Organization During Operations & Maintenance Phase
Figure 2.10 D

For more information on configuration management and the system life cycle, see chapter 7.

Traditionally, CM is linked to managing change in software development. Hence, the largest number of transportation agencies used CM with the software elements of their TMSs. As seen in figure 2.11, however, many of the agencies also used CM to manage change in the following subsections of their TMSs: computer hardware, field equipment, databases, and communication systems.

Figure 2.11 – Subsystems Covered by CM
Figure 2.11 D

The CM process generally consists of the following basic activities: configuration identification, change control, and status accounting and auditing. Some departments surveyed included all of these activities, while others included just a few. Figure 2.12 illustrates which activities were most frequently included in CM processes. Note that the totals in this figure are out of the 19 respondents that indicated they use some sort of formal CM process.

For more information on configuration management processes, see chapter 3.
Figure 2.12 – Agencies Including CM Elements
Figure 2.12 D

CM Organizational Issues

As seen in figures 2.7 – 2.10, transportation agencies are centrally involved in CM through all phases of TMSs' life cycles. Even during the phases in which agencies generally choose to utilize a consultant or contractor for CM support, the agency remains ultimately responsible for changes in the system. This section of the survey addressed two key organizational issues, the use of change control boards and training.

Detailed information about change control boards may be found in the change control section of chapter 3.

Eight of the agencies surveyed used formal change control boards to oversee CM activities. The boards ranged from 1 to 16 people, with an average of 5 people. Most boards met a couple of times a month and also during any emergency situations.

A surprising finding from the survey is that only 7 out of 29 (24 percent) of the individuals responsible for CM had actually received formal training in the area. Those that did receive training most often obtained it in a short-course format. This finding points to the need to provide better CM training opportunities to support the transportation engineering community.

Benefits/Costs of CM

Most of the agencies responding to the survey reported that the benefits gained from CM were well worth the costs required. Table 2.4 presents the average survey rating for a series of CM benefits. The ratings were on a scale of 0 to 10, with 0 representing no benefit and 10 representing the highest level of benefit. Note that according to the survey responses, the largest benefits of CM are seen in the ability to maintain systems and in improved system reliability.

Table 2.4 – Average Benefits Ratings for CM (Scale: 0 – 10)
System Reliability System Maintainability Ability to Upgrade System Ability to Expand System Ability to Share Information with Other Systems Ability to Integrate with Other Systems
7.8 8.3 7.5 7.4 5.8 5.7

Survey respondents also rated various CM costs on a scale of 0 to 10, with 10 being the highest cost and 0 being no cost at all. An important result in this section of the survey is that none of the cost categories received an average score greater than 4.1 out of 10, which indicates that most of the agencies using CM find the costs associated with it to be reasonable. The areas that require the greatest levels of resources are agency personnel time requirements and consultant contract costs. The complete results are displayed in table 2.5.

Table 2.5 – Average Cost Ratings for CM (Scale 0 – 10)
Agency Personnel Time Requirements Consultant Contract Costs CM Tool License Fee Training Costs Lost Productivity due to CM Overhead
3.9 4.1 1.9 2.8 2.4

Finally, when asked to rate if the overall benefits of CM were well worth the costs, on a scale of 0 to 10 (with 10 being complete agreement, and 0 being complete disagreement), 77 percent of the agencies gave a rating of 7 or higher. Again, this strongly indicates that of the relatively small percentage of agencies using CM, the experience has been positive.

CM Benefit Testimonials

Some of the most important information gathered from the survey process can be found in the testimonials of agency personnel on their experiences with CM. Most of the individuals were strong supporters of CM as evident in the following statements.

With almost 20 years experience in the design, implementation, modification and expansion of our system, the benefits of quickly being able to recover from problems by returning to an earlier working state are enormous. Our system has been very dynamic, and there is always some area where we are working on an improvement or upgrade, while still actively managing traffic.

As in any large, complex system, CM can provide a constant understanding of the current state of the system…. The key factor in CM is having a central repository of information for reference as personnel changes occur over the life of the system. It also is a great aid in maintaining the system when items are replaced for repair. Technicians should have ready access to configuration data when installing or re-installing standard system components.

A formal, documented configuration control process can save operational costs over the life of the contract and mitigate the impact of personnel and equipment changes.

Finally, a number of excellent insights into the challenges of instituting CM in a TMSs organization were offered by the Georgia Department of Transportation:

  • User acceptance is slow - people have to become convinced of the importance of CM over time.
  • Development and implementation of CM requires a significant investment in both human resources and capital.
  • CM must be implemented as early as practical in the development of the system and continued throughout the system's life cycle.
  • There is a delicate balance between the time spent on CM and the rewards to be gained.
U.S. Department of Transportation Federal Highway Administration