Integrated Corridor Management Initiative – ICMS Surveillance and Detection Needs Analysis for the Transit Data Gap

3.0 Results of Overall Surveillance and Detection Needs Analysis from Phase 1 of the ICM Initiative

The analysis of the overall surveillance and detection needs for the ICMS operational concepts started with a review of the operational concepts, specifications, and training documents for the ICM Initiative. These include:

1. The ICMS Concept of Operations for a Generic Corridor [31]
2. The ICMS Foundational Research on Corridor Management Strategies [20]
3. The ICMS Surveillance and Detection Needs Analysis [29]
4. The ICMS Concept of Operations documents from each pioneer site [7, 10, 23, 26, 37, 40, 43]
5. The ICMS System Requirements Specifications from each pioneer site [6, 11, 22, 27, 36, 39, 44]
6. The Traffic Control Systems Handbook [15]

3.1    Surveillance and Detection Needs

Appendix C includes twenty need statements that reflect the general set of needs for an ICMS based on this review. The following needs represent the key elements of an ICMS deployment that are impacted by gaps in the transit data:

1. Need to understand demand for transportation services (1.2)
   • Need for corridor performance measures (1.2.1)
   • Need for impact assessment tools (1.2.2)
     • Need to collect information about performance and response of the transportation network (1.2.2.1)

These generic needs point to the kinds of surveillance and detection considered necessary for the corridor management activities. Data is needed to measure or calculate performance measures for the transportation services, and data is needed for modeling the transportation services to help operators understand how the transportation systems will respond to the control actions they may undertake.

Appendix D includes thirty-one detailed needs that were identified in the ICMS Surveillance and Detection Needs Analysis [29]. These detailed needs were summarized as:

1. Needs related to general ICM characteristics
2. Needs related to ICM approaches
3. Needs related to ICM strategies
4. Needs related to ICM operational data

Analysis of the needs, current methods, and typical data sources indicates that surveillance and detection data must support calculation of current performance of a transportation mode and comparison with the design or ideal performance of the transportation mode being monitored.

3.2    Current Surveillance and Detection Capabilities

Surveillance and detection measurements for individual transportation modes are generally based on the control needs for managing the systems without regard to impact on other transportation modes. Additional data is collected based on requirements for reporting to local, state, or Federal transportation agencies. The following values are typically monitored:

Freeway/Tollway Monitoring:

1. Road segment speed (average vehicle distance traveled/time unit) – sampled at 30-120 second intervals, with date/time stamp
2. Road segment volume (vehicles/time unit) – sampled at 30-120 second intervals, with date/time stamp
3. Road segment occupancy (% of unit length lane occupied by vehicles) – sampled at 30-120 second intervals, with date/time stamp

Transit Monitoring:

1. Volume (passengers/route leg) – by time of day and day of week, sampled at each stop, but reported at end of day
2. Fare collected/route leg – by time of day and day of week, sampled at each stop, but reported at end of day
3. Schedule adherence (difference between vehicle actual arrival/departure vs. scheduled arrival/departure) – current and daily summary/route, sampled at 30-300 second intervals, with date/time stamp
4. Vehicle location (with varying degrees of accuracy and update times), sampled at 30-300 second intervals, with date/time stamp

Parking Management Monitoring:

1. Volume (number of vehicles using the parking facility) – current and daily total, sampled as vehicles enter or leave facility, reported daily as either hourly or daily volume
2. Parking spaces remaining – current, sampled as vehicles enter or leave facility, reported daily as hourly and daily counts. In some cases, the current data is displayed at the entrance to the lot.

Arterial Monitoring:

1. Call (vehicle/pedestrian presence) – sampled 0.1 to 0.01 seconds, rarely reported to a central server
2. Volume (number of vehicles passing a point on the roadway during a specified time period) – sampled 0.1 to 0.01 seconds, typically reported as 5 minute summary and archived as hourly totals by time of day
3. Road segment occupancy (percent of time that a point on the roadway is occupied by a vehicle) – sampled 0.1 to 0.01 seconds, most systems report current average per unit time
4. Road segment speed (distance traveled by a vehicle per unit time) – sampled 0.1 to 0.01 seconds, typically reported as 5 minute average
5. Queue length (number of vehicles stopped in a lane behind the stop line at a traffic signal) – sampled 0.1 to 0.01 seconds, typically reported as 5 minute average
6. Headway (time difference between beginning of successive vehicle detections) – sampled 0.1 to 0.01 seconds, typically reported as 5 minute average

It should be noted that in the above list of data monitored, the performance measures that are reported are not generally the values that are used to manage the performance of the transportation modes. For example:  volumes are reported on highways, but speed and occupancy are the values used for responsive ramp metering. Passenger volume is reported on transit systems, but current schedule adherence values are the measurements used to control transit signal priority and to make real-time decisions about schedule and route deviations. Daily volume is reported for parking facilities, but signs and access controls are driven off of the number of spaces remaining. Arterial reporting is primarily based on volume and level of service (speed), but local signal controls use call and queue length for the primary control parameters.

It is significant that the latency (time between data sampling and data reporting) differs substantially across current data collection systems. The difference ranges from seconds on freeway networks, to minutes on arterial networks, to daily in the case of some transit network and parking lot data.

It is problematic that some of the data is reported without time/date stamps that would permit the data to be aligned with other information sources for analysis. Problems are also introduced where observations are sampled at one rate and reported as “rolled-up” averages or totals over much longer periods of time. While this practice may save communication or storage costs, it considerably limits the usefulness of the resulting data.