Lessons Learned: Monitoring Highway Congestion and Reliability Using Archived Traffic Detector Data
4.0 Next Steps
A high level of data quality is absolutely essential for an archive to be useful to a wide variety of interests (including performance measurement. If users perceive that the data are not of sufficient quality, the archive will not get used and interest will wane. The best way to ensure quality data is to have the original collectors (owners) take responsibility for it. This includes developing formal review procedures (which can be automated through software), routinely publishing data quality statistics, and establishing a feedback process whereby users can alert collectors/owners of quality problems not originally detected. However, that is the easy part. A much more difficult part of maintaining a high level of data quality is ensuring that field devices are properly installed, calibrated, and maintained. These activities require significant investment by data collectors/owners.
Actions: It would be useful to document the costs of proper detector installation, calibration, and maintenance activities, especially with regard to type of equipment and the level of data quality (accuracy in the field measurements) achieved. Identifying best practices for each of these activities would also foster archive development and use. Promoting the use of quality control software by data collectors/owners (i.e., TMCs) would also support maintenance of quality data.
4.2 Presentation and Use of Congestion Performance Measures by Transportation Agencies in Decision-Making
As discussed earlier, identification of which performance metrics (measures) should be used in a congestion monitoring program has received a good deal of attention over the past few years. The remaining three pieces of the performance measurement process are: What data are needed, how should the measures be presented, and how should the measures be used in the decision-making process. This next step deals with the second and third of these issues. The toughest of these two issues is how performance measures influence investment and policy decisions.
Actions: (1) A scan should be conducted of different methods being used by transportation agencies to present congestion performance measures to the public and decision-makers. From this, a compendium highlighting the most effective presentation methods should be compiled. (2) Case studies of two or three transportation agencies that have aggressive congestion performance monitoring programs in place should be conducted to document how the measures have influenced investment and policy decisions.
The congestion performance measures developed so far focus mainly on an overall picture of congestion using traffic detector, probe, or modeled data. However, to be more useful for implementing operations strategies, the causes of congestion should be tracked at a detailed level. In other words, what factors ("events") have contributed to overall mobility and what are their magnitude; factors include traffic incidents, weather, work zones, changes in traffic demand, special events, and recurring bottlenecks. If the share of total congestion attributable to these sources can be produced, strategies targeted at the root causes can be developed. Identifying the events that are restricting mobility is important at both the national level (development of overall programs) and the local level (development of specific actions). Key in this effort is the capture of roadway event-related data in a consistent manner. These data must be fully integrated with traffic detector and other forms of traffic data so that the events' influence on congestion patterns can be ascertained.
Actions: (1) An effort should be undertaken to harmonize the data requirements required for documenting roadway events from performance measurement and archive perspectives as opposed to purely an operational perspective. This involves review of and potential modification to existing ITS standards (especially the TMDD and 1512 "family") and standards used in the data systems of nontransportation agencies (especially police computer-aided dispatch systems. (2) A scan of current event/traffic data integration practices among ITS data archives would reveal best practices and potential pitfalls.
As shown previously in Figure 2.1, several methods are available for developing congestion performance measures. Many areas must use a mix of these measures depending on the availability of data from deployed ITS. For example, some roadways may have detectors, others may have problem systems, others may have travel times collected by floating cars, and others may only have purely modeled travel times. However, the relationship of all these methods to the measures they produce has not been determined. For example, how compatible are estimates from detectors versus those from probes versus different models - are they reasonably close or wildly different? Are there ways to adjust one method to match another?
Actions: A study utilizing data from a variety of sources within 1-2 metropolitan areas should be explored. Original data collection to establish a baseline ("ground truth") may be required. Data should include roadway detectors, vehicle probes, and transportation models at a minimum.
Self-assessments have been used successfully in the emerging fields of operations. Self-assessments are essentially expanded checklists that agencies can use to compare their current activities to an accepted practice. They are very useful in allowing agencies to identify areas that they have not considered; once identified, they can seek out additional guidance.
Actions: It is recommended that an operations performance measurement self-assessment process be developed for use by State and local transportation agencies. The self-assessment can be based on much of the work presented in this report and should include several features:
- Identification of good, better, best practices in monitoring the performance of transportation systems from an operations perspective. Monitoring should include:
- System performance from the user's perspective ("outcome" congestion/mobility metrics based on travel time)
- System performance from the agency's perspective ("output" metrics such as the incident timeline; work zone activities and durations)
- Emergency preparedness
- Institutional relationships required for data collection and for improving field activities
- Measurement methods/models and data collection programs