Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

10 Measuring and assessing rural DRT performance require that performance data be identified and defined. This is an important step that will help improve consistency in data definitions and report- ing. Particularly at rural transit systems there may be limited administrative staff for data collection tasks, and these tasks may be burdensome without technology tools such as a computer-assisted scheduling/dispatch (CASD) system. Data on DRT may be intermingled with data for other service modes, such as route deviation, but consistent data reporting practices are needed for assessing rural DRT performance. This chapter identifies key performance data for rural DRT systems and also reviews other data elements that are often collected for performance assessment purposes. 3.1 Performance Data—Which Data Elements Are Particularly Important? This Guidebook’s companion guidebook on urban DRT systems, TCRP Report 124, identi- fied a long list of data elements that DRT systems can consider for data collection and assessment purposes. This list was then distilled to six key data elements for performance assessment, in keeping with the research project’s objective of selecting a limited number of performance data elements and measures. For rural DRT systems, a similar set of six data elements is used for this Guidebook as the key data for performance assessment purposes: • Vehicle-hours, • Vehicle-miles, • Passenger trips, • Total operating expense, • Accidents/safety incidents, and • On-time trips In addition to these data elements, a number of others are identified and discussed in this chapter. These additional data elements and related performance measures provide rural DRT systems with additional resources for assessing their performance. 3.2 Performance Data for Rural DRT—Now There Is NTD With the Rural NTD established in 2006 for transit systems operating in rural areas, transit managers at the nation’s rural systems must now comply with specific federal reporting require- ments. NTD reporting has been a staple for urban transit systems since the 1970s, and adoption of standardized reporting for rural systems in 2006 is a positive progression. C H A P T E R 3 Performance Data for Rural DRT

The NTD reporting requirements can help standardize data, data definitions, and data reporting practices for rural systems. Over time, the national database will provide a large reservoir of rural transit information that can be useful in assessing performance trends over time and providing peer information that an individual transit system can use for reviewing its own performance. However, it should be noted that some aspects of the federal transit report- ing system have evolved over the years, and in some cases, continue to evolve. Take the NTD definitions related to safety, for example. These data elements have been revised several times over the years for urban reporters. Such is the case for Rural NTD as well, even over its short history. Certain data elements included in its 2006 inaugural year have been revised since then. For this Guidebook, some of the NTD definitions are used and, in other cases, not, as is discussed in the following section. 3.3 Key Performance Data for Rural DRT Performance Assessment The key performance data used for this Guidebook and their definitions are identified below. Vehicle-Hours Vehicle-hours measure the time from when the DRT vehicle leaves the garage (or other starting location) to go into service until the time that the DRT vehicle pulls in after completing service— “pull-out to pull-in.” Vehicle-hours therefore include deadhead time needed to travel to the first pick-up location and from the last drop-off location back to the garage. Vehicle-hours do not include any charter service, vehicle operator training, or vehicle maintenance testing. Vehicle- hours also do not include any scheduled time off for the operator such as a formal lunch break. While the urban NTD reporting program requires transit agencies to report vehicle-hours (as well as vehicle revenue hours), Rural NTD does not ask for “vehicle-hours.” Rather, Rural NTD asks systems to report vehicle revenue hours. The standard definition for vehicle revenue hours is the time from first passenger pick-up to last passenger drop-off, not including any deadhead time, or any charter service, vehicle operator training, vehicle maintenance testing, or scheduled operator time off such as a formal lunch break. However, somewhat confusingly, the NTD’s 2008 Rural Reporting Manual (the current man- ual at the time this Guidebook was developed) defines vehicle revenue hours for the DRT mode (and only for the DRT mode) as the time from when the vehicles pull-out to go into revenue ser- vice to the time they pull-in from revenue service. In other words, vehicle revenue hours for rural DRT are defined the same way as vehicle-hours are typically defined, which is pull-out to pull-in. For this Guidebook, it was determined that the appropriate data element for rural DRT would be vehicle-hours, and it is defined as “pull-out to pull-in,” minus any scheduled time off for the operator. Data Collection for Vehicle-Hours Vehicle-hours data are obtained from vehicle operator logs. These logs, also called “manifests,” should be configured so that the operators report the actual times that they leave the garage or other starting location (e.g., their home if they take the vehicle home with them at night) and the time that they return at the end of their driving shift, referred to as “pull-in.” Performance Data for Rural DRT 11

The log should also be configured so that the operators report their scheduled time off, such as lunch breaks, both the starting and ending time of their break (if they have such a break), or any other time that they are formally not providing or available to provide transportation. This does not include the time it might take for an operator to quickly grab something to eat when traveling between pick-up and drop-off points. Any scheduled time off, when the operator is not providing or available to provide transporta- tion, needs to be deducted to determine total vehicle-hours. For example, an operator may have a driving assignment in the morning hours and then once back at the garage, that operator is assigned other duties in the office such as answering the phone and taking trip reservations before another driving assignment in the afternoon. The logs must be designed so that these data are collected, and reporting practices must ensure that this time that the operator spends in the office is not included in vehicle-hours. Vehicle-hours, then, include all the time from pull-out to pull-in, minus scheduled time off or time that the operator is assigned to other, non-driving duties. Vehicle-hours do not cover transportation activities such as exclusive school bus service and charter service. If volunteers are used to provide DRT service, their time is counted as vehicle-hours, using the same definition—the time from garage (or other storage location) pull-out to garage pull- in, exclusive of any scheduled time off. If a volunteer driver accompanies the passenger to an appointment as part of the trip (e.g., a medical appointment), the time spent at the appointment is counted in the same way as scheduled time off; in other words, it is not counted as part of vehicle-hours. Where volunteers are used, the DRT system’s reporting forms should be devel- oped so that data on the volunteers’ vehicle-hours are captured. These hours would then be added to the vehicle-hours provided by paid drivers. If taxis are used to provide some of the rural transit system’s demand-response service, the time that the taxi is transporting the system’s passengers should also be included in the DRT system’s count of vehicle-hours. The agreement with the taxi provider should include require- ments that the company provide data showing the elapsed time for the trips provided per the arrangement with the rural transit system. Since taxi companies are more accustomed to deal- ing with miles rather than hours, this may require some emphasis to ensure the time data are provided. Vehicle-Miles Vehicle-miles include the miles from when the DRT vehicle leaves the garage (or other starting location) to go into service until the time that the DRT vehicle pulls in after completing service— “pull-out to pull-in.” Vehicle-miles include deadhead miles needed to travel to the first pick-up location and from the last drop-off location, but do not include any miles for charter service, vehicle operator training, or vehicle maintenance testing. Vehicle-miles, therefore, correspond to vehicle-hours; they are the miles traveled while the DRT vehicle is accumulating vehicle-hours. Vehicle-miles should also include the miles of any personal vehicles used in DRT service, such as by volunteers who use their own cars, and the miles of any taxicabs that provide DRT service for the transit agency. It is noted that Rural NTD does not ask for vehicle-mile data, but rather vehicle revenue miles, defined by Rural NTD as the miles traveled by the vehicles when in revenue service. Data Collection for Vehicle-Miles Vehicle-miles data are obtained from vehicle operator logs. These logs should be configured so that the operators report odometer readings when leaving the garage or other starting loca- 12 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance

tion (e.g., their home if they take the vehicle home with them at night) and the time that they return at the end of their driving shift. The log should also be configured so that any miles operated during formal lunch breaks or any other time that they are formally not providing or available to provide transportation are tracked. These miles, traveled when the operator is not providing or available to provide trans- portation, need to be deducted to determine total vehicle-miles. Vehicle-miles data do not include any miles traveled for exclusive school bus service or charter service. If volunteers are used to provide DRT service, their distances traveled are counted as vehicle- miles, using the same definition—the miles traveled from garage (or other storage location) pull- out to garage pull-in, exclusive of any scheduled time off. If taxis are used to provide some of the rural transit system’s demand-response service, the miles that the taxis accumulate when transporting the system’s passengers should also be included in the count of vehicle-miles. The agreement with the taxi provider should include requirements that the company provide data on the miles of all trips provided on behalf of the DRT system as well as time. Potential Issues When Reporting Vehicle-Hours and Vehicle-Miles Regarding the reporting of hours and miles data, there are several possible errors that may be made by transit systems. These include the following: • Use of operator pay hours for revenue or vehicle-hour data. Some DRT systems mistakenly compute vehicle-hour or revenue hour data from operator time records—that is, they report the time as the sum of operator pay hours. This is not correct. For vehicle-hours, the time includes only the time from vehicle pull-out to vehicle pull-in, minus scheduled breaks, and it does not include such time as that provided to operators for pre-trip or post-trip inspections. • Estimation of hours and miles data. NTD requires the direct recording of hours and miles data. Experience with urban NTD reporting has found that some reporters incorrectly assume that estimated values can be reported, but this is not the case. • Incorrect treatment of lunches and breaks. Some transit systems do not subtract out scheduled lunches and breaks from vehicle-hours; therefore, these systems overstate their hours data. • Incorrect calculation of total vehicle-miles. Transit agencies may calculate total vehicle- miles by simply subtracting their vehicles’ starting odometer reading at the beginning of the year from the odometer readings at the end of the year. This is not correct and leads to an over-reporting of vehicle-miles. By using year-start and year-end odometer readings, the mileage figure will include such miles as those needed for vehicle operator training, main- tenance testing and other maintenance needs, and possibly charter or other purposes. These latter miles are not included for NTD reporting purposes. Passenger Trips The data element passenger trips is also called ridership. NTD uses the term unlinked passenger trips or regular unlinked passenger trips according to the Rural NTD manual. The Rural NTD has added an additional term for passenger trips: sponsored unlinked passenger trips, which is defined as DRT trips that are paid in part or in whole by a third-party such as a human service agency or Medicaid agency. This Guidebook uses the term passenger trips. The data element is a count of the number of passengers who board the DRT vehicle, with passengers counted each time they board a DRT vehi- cle. If a passenger travels with children or a Personal Care Attendant (PCA), these other riders are also counted as passenger trips. Total passenger trips is the sum of what Rural NTD refers to as regular unlinked passenger trips and sponsored unlinked passenger trips. Performance Data for Rural DRT 13

Passenger trips should also include the count of any passenger trips provided by taxicabs if taxis are used for the DRT service. Additionally, passenger trips provided by volunteers should be included in the total DRT passenger trip count. Data Collection for Passenger Trips Passenger trip data are obtained from driver logs, which should be designed so that vehicle operators record the number of passengers boarding at each pick-up location. For volunteer services, reporting forms should be developed that will capture passenger counts. PCAs and companions are counted as passengers on DRT as long as they are not employees of the transit system. Attendants and companions are included regardless of whether they are fare-paying passengers. Transit system employees are not counted as passengers if they are performing work duties that require traveling on the vehicles and are being paid while traveling. Examples of these work duties include observing vehicle operations and serving as an on-board aide or assistant for the passengers. However, transit system employees are counted as passengers if they are traveling for personal reasons including commuting to and from work. There can be inconsistent reporting of passenger trips by DRT systems, with common problems being • Counting only round trips. Some DRT systems only count round trips of passengers in their total count of passenger trips, so when a rider is transported, for example, to the senior center and then back home, the transit system might count one passenger trip. This is not correct. That rider should be counted each time she boards the vehicle; in this example, two passenger trips should be counted. • Incorrect definition of passengers. Some DRT systems incorrectly define passengers in one of the following manners: all fare-paying passengers, ADA-certified riders only, all persons boarding not including children, and all persons boarding not including PCAs and com- panions. This also is not correct. Each person who boards the vehicle, including someone who rides for free, a PCA, a companion, or a child, is defined as a passenger. Total Operating Expense Total operating expenses or total operating costs include those costs needed to operate and administer transit services day to day. These costs include salaries and wages, fringe benefits, materials and supplies, insurance, taxes, and outside services such as cleaning and utilities. These are the costs used for the day-to-day expense of operating and maintaining vehicles; maintain- ing other equipment, buildings and grounds; and general administration costs including mar- keting and customer support, finance and procurement, planning and service development, and legal costs. Total operating costs also include equipment lease and rental costs. Operating costs include depreciation of capital and the interest paid on loans for capital purchases. Importantly, operating costs exclude the purchase of capital such as vehicles, equipment, or facilities. As defined by Rural NTD, operating costs include the costs to operate “sponsored” transit pro- grams and services such as those provided for client-specific or categorical programs—for exam- ple, Meals-on-Wheels, sheltered workshops, independent living centers, and any social service agency programs. The Rural NTD, however, does not ask reporters for operating costs by mode. Rather, Rural NTD asks for the annual operating expenses for all the rural transit services oper- ated, including DRT. Thus, rural transit systems do not need to allocate costs by mode for federal reporting purposes. 14 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance

For purposes of this Guidebook and performance assessment, however, it is important that the operating costs be separated by transit mode, when the rural agency operates more than just DRT. Most transit expenses are known as direct expenses and can be associated on a one-on-one basis with a specific transit mode. Examples include operator labor and fuel costs. However, some expenses, known as indirect or shared expenses, cannot be directly associated since they support several modes. The majority of these costs, often called “overhead costs,” are adminis- trative such as accounting and planning. These costs must be allocated on a reasonable basis to the individual transit modes. There are a number of transit references that can be consulted on cost allocation and these are listed in References in this Guidebook. Data Collection for Total Operating Costs The transit system’s accounting system records and reports operating expenses. It is impor- tant that all operating expenses are included. For example, if the rural DRT system contracts for service and the contractor obtains fuel from the city or county yard, the costs for the fuel should be included as part of operating costs. If the transit system purchases transit services and the contractor provides vehicles, it needs to ensure that the contractor provides cost data that are separated into operating and capital costs. This requirement should be explicitly stated in the contract to make it clear to the operator before service starts. Where there are problems with DRT system reporting of operating expenses, it is often related to two issues: 1. Poor cost allocation of joint expenses. Transit systems that operate more than one mode or type of service must allocate certain shared expenses such as the administrative costs and shared building expenses. Some transit systems may not use reasonable allocation pro- cedures or use procedures that should be updated. 2. Inclusion of capital charges in purchased transportation costs. While contracting for service is less common for rural DRT systems than for urban DRT, there may be cases where the DRT service contractor provides the vehicles. The contractor may charge one unit cost per hour or per trip that includes both the operating costs associated with vehi- cle operations and maintenance and the capital cost associated with providing the vehicle. Where this is the case, the transit system should require the contractor to report in its monthly invoice the portions of the invoiced costs that are capital and operating so that operating costs can be accurately reported. Accidents Data on accidents and other safety-related incidents are important for all transit systems to collect and assess. Transit systems may have different procedures related to how accidents and safety incidents are formally reported, and most systems differentiate between preventable and non-preventable accidents. For purposes of this Guidebook as well as for TCRP Report 124, it was decided that the NTD definitions should be used for safety data since they provide standard definitions. Rural NTD requires the collection and reporting of three types of safety data: reportable inci- dents, fatalities, and injuries. Per NTD, these are defined as follows: 1. Reportable incident: The existence of one or more of the following conditions constitutes a reportable incident: a fatality, injuries requiring immediate medical attention away from the scene for one or more persons, or property damage equal to or exceeding $25,000. A Performance Data for Rural DRT 15

reportable incident must be related to the operation of revenue service and not associated with unrelated tasks. The following types of incidents are not reportable to the NTD: – Mechanical, – Industrial, or – Administrative work orders. 2. Fatality: A transit-caused death confirmed within 30 days of a transit incident. 3. Injury: The definition of injury requires immediate medical attention away from the scene of the accident. This means that injured people who receive treatment at the accident scene are not counted toward the definition of an injury. It also means that injured people who delay receiving treatment after the accident are also not counted. Data Collection for NTD Safety Data The data on accidents is collected as part of the claims management function. Specific reports should be prepared on a routine basis to document the events that meet the accident criteria. Since DRT systems typically collect data on all safety-related incidents and accidents regardless of NTD reporting thresholds, DRT managers have essentially two sets of data on safety incidents. It is important to note that the NTD definition is event-oriented and goes beyond measuring injuries to passengers and includes events related to property damage. Since each reportable safety incident is an event, there can be more than one “result” from that event. For example, there may be an accident with property damage exceeding the dollar threshold and also with three passenger injuries meeting the NTD definition of injury. This would be counted as one reportable incident, and additionally there would be three injuries reported. There may be an accident with limited property damage yet with two passenger injuries meeting the reporting def- initions. As with the first example, one reportable incident would be tallied, and additionally two injuries would be reported. Several issues may be problematic with collection of the NTD safety data: • Definition of injury. According to NTD, an injury requires immediate medical attention away from the accident scene. Immediate medical attention includes, but is not limited to, transport to the hospital by ambulance. If an individual is transported immediately from the incident scene to a hospital or physician’s office by another type of emergency vehicle, by passenger vehicle, or through other means of transport, this is also considered an injury. An individual seeking medical care several hours after an incident or in the days following an incident is not considered to have received immediate medical attention. In cases that are less clear-cut, reporters can apply their judgment in determining whether the injury sustained caused the individual to immediately seek medical attention. The medical attention received must be at a location other than the location at which the incident happened. The purpose is to exclude incidents that only require minor first aid or assistance at the scene, but this distinction is not supposed to be burdensome for the transit provider. The NTD does not require the system to follow-up with each person to see if they received medical attention at the hospital. • Definition of fatality. A fatality is a transit-caused death that occurs within 30 days of a transit incident. If death occurs after 30 days, it is classified as an injury. • Accident must involve a transit vehicle or happen on transit-controlled property. This means that the accident occurs in an environment under the direct control of the transit system. For DRT systems, this definition typically limits the counting of accidents to those involving transit vehicles since most DRT systems do not own or control other transit facil- ities used by passengers such as stations, buildings, or shelters. 16 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance

On-Time Trips On-time trips are those trips where the DRT vehicle arrives at the scheduled pick-up location within the DRT system’s definition of on-time. Some rural DRT systems measure their on-time performance at the drop-off end, so an on-time trip is one that gets the rider to his or her desti- nation by the promised time. Often, DRT systems provide a “window of time” within which the vehicle’s arrival is defined as on-time. A 30-min window at the pick-up end is typical for many urban and some rural DRT as well, but variations are used, including windows from 15 to 60 min. Some rural systems determine that they are on-time if the vehicle arrives anytime before the scheduled pick-up time. Thus, there is no window of time—any time before the scheduled pick- up time is on-time. However, while many rural DRT provide their riders either a specific pick-up time or a window of time for the pick-up and many have their operators report arrival times at scheduled locations on their manifests, the majority do not use the data formally or routinely to report on-time performance. This is a data element, however, that is important for performance- assessment purposes. Trip timeliness is perhaps the single most important measure of service qual- ity from the DRT rider’s perspective and provides an assessment of the DRT system’s reliability. Data Collection for On-Time Trips While data collection and reporting of on-time trips may not be routinely done by rural DRT systems, it need not be burdensome. DRT systems should ensure that operators record on their manifests the time that they arrive at the pick-up location (not the time the rider boards, but the time when they have pulled up to the scheduled location). They should also record the time when they arrive at the drop-off location (again, the time once they have pulled up at the destination). Alternatively, some DRT systems have their operators call in on the radio their arrival and drop-off times. In such cases, the dispatcher maintains the records and should ensure that the data are reported on the schedules so that comparisons can be made of the actual times versus the scheduled times. For those rural DRT systems that have Mobile Data Terminals (MDTs) in their vehicles, the pick-up and drop-off times can be reported electronically, via MDT, which greatly eases the data reporting effort. If the transit system has Automatic Vehicle Location (AVL) technology, this will enable the system to “see” where the vehicles are, and operator-reported data on trip timeliness can be verified. There are several issues with the reporting of on-time trips from a data reporting perspective, including • The definition of on-time varies. Since there are different ways that on-time is defined, reported data on timeliness have somewhat different meanings. If a rural system defines on-time using a 45-min window and 90% of its trips are measured as on-time, can this be compared with a system with a 30-min on-time window and 90% on-time trips? Not really. What is more important is that the rural DRT systems have defined on-time and that they measure timeliness on a routine basis, allowing internal comparisons over time. • On time at the pick-up or the drop-off end. While DRT systems typically focus on mea- suring on-time performance at the pick-up end, timely service at the drop-off end may be more important for the passenger, particularly for a time-sensitive trip such as to work or a medical appointment. It is recommended that DRT systems monitor on-time perfor- mance at the drop-off end for those trips that are time sensitive. It is important to note that DRT systems need to schedule riders’ trips based on either the desired pick-up time or the desired drop-off time; a DRT system cannot schedule a rider’s trip with both a requested pick-up time and requested drop-off time, given that both com- Performance Data for Rural DRT 17

puterized and manual scheduling procedures must have some degree of flexibility for sched- uling purposes. Riders must choose which end of the trip to focus on when requesting their trips. When riders have time-sensitive trips, the DRT system is better able to provide a timely arrival when riders accept a DRT system-determined pick-up time that is based on meeting the appointment time. Otherwise, riders may face late arrivals for appointments as they may not allow adequate time for DRT shared-ride service. • Self-reporting bias. For most rural DRT systems, the on-time data are recorded by vehicle operators on their trip manifests. There may be tendency for some operators to “round” the pick-up times to better fit within the on-time window or to make their arrivals seem timelier. Use of MDTs/AVL can help with the reporting of accurate data. The AVL data can be used to verify operators’ locations at specified times, providing a check on operator reporting and specifically on-time performance data. Some DRT systems have been sur- prised to learn their “true” on-time performance once they have transitioned from opera- tor reporting via manifest to MDTs/AVL. Differences of more than 5 to 10 percentage points for on-time performance are not uncommon once a DRT system transitions from manual reporting to MDTs/AVL. 3.4 Other Performance Data for Rural DRT Performance Assessment In addition to the six data elements discussed above, there are other data elements that are useful for DRT systems to collect and report. These are identified below. Revenue-Hours While the Rural NTD does not require the collection or reporting of revenue-hours for rural DRT systems,1 this is a useful data element that some rural DRT systems already collect. Revenue- hours are defined as all time from the point of the first passenger pick-up to the last passenger drop-off, as long as the DRT vehicle does not return to the dispatching point (the transit agency’s garage or other designated location where the operator is waiting for a passenger trip assign- ment), minus scheduled time off such as driver lunch breaks. It is noted that, for DRT service, there may be periods of time when there are no passengers riding or when the vehicle is stopped before proceeding to the next pick-up. This time is consid- ered revenue time as long as the vehicle operator does not return to the dispatching point. Sched- uled time off is typically a lunch break, but this does not include the time it might take for an operator to quickly grab something to eat when traveling between pick-up and drop-off points. Data Collection for Revenue Hours Revenue-hours data are obtained from vehicle operator logs. These logs, also called “mani- fests,” should be configured so that the operators report the actual times that they go into and out of revenue service. This is in addition to reporting the times that they leave the garage or other starting location (e.g., their home if they take the vehicle home with them at night) and the time that they return at the end of their driving shift, referred to as pull-in. The difference between vehicle-hours and revenue-hours is deadhead time. 18 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance 1 As explained earlier in this chapter, Rural NTD asks reporters to report “vehicle revenue hours,” but defines this data element, for the DRT mode only, as the time the vehicles pull-out to go into revenue service to the time they pull-in from revenue service, which is the definition typically used for vehicle-hours, not revenue-hours.

The log should also be configured so that the operators report their scheduled time off such as lunch breaks, both the starting and ending time of their break (if they have such a break), or any other time that they are formally not providing or available to provide transportation. Scheduled Trips Scheduled trips are those trips that are placed onto vehicle schedules for transportation service. Some DRT systems accept trip reservations for trips up to 2 weeks, and sometimes even longer, in advance. Others have a shorter advance reservation time period, accepting trips no more than 1 week or several days in advance. If the DRT system provides ADA paratransit service, it must accept trip reservations on a next-day basis. That is, an ADA-eligible rider should be able to make a trip reservation for tomorrow up to the close of the DRT system’s office hours today. Some DRT systems provide immediate response service, accepting trips from riders on the same day of desired travel or even just 1 hour or less before a desired trip. Those DRT systems that are advance reservation will often accept a same-day trip request for an urgent trip or when they have available capacity. All of these trips are scheduled trips. Data Collection for Scheduled Trips This data element is collected by the DRT system call-takers and schedulers. For those systems that have a CASD system, this data element is calculated automatically by the system. Completed Trips Completed trips are those trips where a passenger is transported from an origin to a destina- tion. A completed trip may carry more than one passenger if there are two or more passengers traveling from the same origin to the same destination. Completed trips are a subset of sched- uled trips, as some trips will be cancelled and some (hopefully not too many!) will be no-shows. Data Collection for Completed Trips Data on completed trips are collected by vehicle operators, who report the data on their man- ifests. Some DRT systems that operate as immediate response may have the operators call in their passenger counts via the radio as the passengers are picked up and dropped off—“plus one at 1429 Elm St. . . . minus one at 2021 Elm St.,” and so forth. In this case, the count of completed trips is handled by the dispatchers. Or for the rural DRT systems that have MDTs and a CASD system, passenger data can be collected electronically. No-Shows A no-show is defined as a failure of a rider to show up for a scheduled trip at the scheduled time and location, when the vehicle has arrived in a timely manner (according to the DRT sys- tem’s definition of on-time), and when the rider has not cancelled the trip in advance. Should a rider cancel the trip when the vehicle arrives (“cancel at the door”) or if a rider cancels the trip “late” (with late cancel defined in different ways by DRT systems), then those, too, are typically counted as no-shows. DRT systems may differentiate between no-shows and late cancels, but since a late cancellation often has the same negative effect on DRT service, the two data elements are usually counted and assessed together. Data Collection for No-Show Trips No-shows are reported by vehicle operators. Late cancels are reported by dispatchers or call- takers. Regarding the reporting of no-shows, some DRT systems require that vehicle operators obtain approval from a dispatcher before marking a passenger a no-show. This allows the dispatcher Performance Data for Rural DRT 19

the opportunity to try and telephone the passenger so that the trip might be provided, and it also provides a level of supervision over the vehicle operator. Late Cancellations A late cancellation is a trip cancellation that is made shortly before the vehicle is scheduled to arrive. The exact time that a cancellation becomes a late cancellation rather than just a cancella- tion is a matter of DRT policy and should take into account the degree to which the scheduling/ dispatch function can “re-use” the space newly created by the cancellation. For a rural DRT sys- tem that must travel long distances between pick-ups and drop-offs, a late cancellation may be one that is made less than 3 or 4 hours before the scheduled trip, or some rural systems may require cancellations the day before the scheduled trip. On the other hand, if the rural system operates predominately within a small community, a late cancellation may be one made with less than 1 hour’s notice. However the DRT system defines a late cancellation, the system must ensure that its riders and the community understand the defi- nition as well as why it is so important to cancel unneeded trips in as timely a manner as possible. For rural DRT systems that provide ADA paratransit service, it should be noted that the FTA has commented that a late cancellation for purposes of rider sanctions should be the operational equivalent of a no-show. Data Collection for Late Cancellations Trip cancellation data are reported by the dispatchers or other DRT system staff that work in the dispatch office. Missed Trips A missed trip is defined as a failure of the vehicle to show up for a scheduled trip. A missed trip can also be defined to include a trip that arrives so late that the passenger is no longer there for the trip or declines to take the trip. In both cases, the trip is not completed. This more expan- sive definition of a missed trip is recommended. Some DRT systems use the term vehicle no-show instead of missed trip. There may also be con- fusion with the term “missed trip” since some DRT systems combine missed trips and late trips together for purposes of contractor monitoring or for other reasons. They are not the same oper- ationally: a missed trip results with an incompleted passenger trip because of vehicle operator or other DRT system error, whereas a late trip, no matter how late, results in a completed trip. Data Collection for Missed Trips Data on missed trips are obtained from vehicle operator records or dispatcher data. Rural DRT systems may not routinely collect data on missed trips, but it should be monitored if missed trips become problematic. Trip Denials A trip denial is a DRT trip that is requested by a passenger, but that the DRT system is not able to provide typically because capacity is not available at the passenger’s requested time. For those rural DRT systems that provide ADA paratransit, the definition of denials is considerably more complicated, and the reader is referred to TCRP Report 124. It can be useful for a DRT system to collect trip denial data as it provides hard evidence of lim- ited capacity, at least during the times when trips are denied. These data may be useful when the 20 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance

DRT system is trying to make a case that added capacity is needed, or it could suggest that sched- ules for vehicle operators should be reviewed to ensure that the operators’ shifts match ridership demand patterns. Data Collection for Trip Denials Data on trip denials are captured during the trip reservation process by call-takers or schedulers. Trip Length Trip length is the distance measured in miles from the passenger pick-up to drop-off location. The cumulative sum of all passengers’ trip lengths is referred to as passenger-miles and is a NTD data element required for urban reporters. While rural transit systems do not have to report trip length data for NTD purposes, it is use- ful data that rural DRT systems may want to assess on a sampled basis. It is particularly useful where a rural system wants to track and monitor the longer distance trips that it may serve and to assess the impacts of those long trips on its performance. In such cases, the DRT system should collect, report, and monitor trip length data. Data Collection for Trip Length One approach would be to have the vehicle operators record odometer readings at each pick- up and drop-off location for a sampled time period—for example, a 1-week period several times per year—which would provide the trip lengths of the sampled passenger trips. Another approach to data collection would be to use an FTA-developed procedure that is designed for urban reporters to meet passenger mile reporting, as described in the FTA Circular C 2710.2A Sampling Procedures for Obtaining Demand Response (DR) System Operating Data Required under the Section 15 Reporting System (available at www.ntdprogram.gov). While these procedures are designed to meet FTA’s defined statistical accuracy levels for urban reporters, a rural system does not have these requirements and could use the procedures for a time period it determined appropriate—for example, a sampled month twice per year. Travel Time Travel time is the time that the passenger spends on-board the vehicle from time of boarding to arrival time at the destination. Travel time is a useful data element, providing data to help measure both the degree to which the scheduling function has grouped similar passenger trips for greater efficiency and service quality from the passengers’ perspective. Travel time is also important to monitor for ADA paratransit systems. Under the regulations concerning capacity constraints, systems must not have substantial numbers of trips with excessive travel times. Data Collection for Travel Time Data on travel time are recorded by vehicle operators through their manifests or through MDTs. The time at both pick-up and drop-off locations is needed to determine travel time. Complaints For a transit system, a complaint is an expression of dissatisfaction by a passenger or the pas- senger’s representative over some aspect of service. Transit systems typically monitor complaints that are related to service and those complaints over which they have day-to-day control. Some transit systems refer to these as “valid complaints.” Transit systems may or may not record com- plaints that are related to matters of policy—for example, the operating days or hours. Nonetheless, Performance Data for Rural DRT 21

such information may be useful over time. If a DRT system repeatedly hears complaints from its riders and the community that there is no weekend service, the transit system may want to con- sider whether it could expand its service days. Data Collection for Complaints Transit system staff can collect complaint information in varying ways: by letter, by telephone call, and via e-mail. Use of all these methods will facilitate passenger feedback. Where the DRT ser- vice is contracted, complaints are typically directed to the public entity sponsor. Complaint data should be summarized on a routine basis and assessed internally rather than compared with other DRT systems, given the range in how complaints are defined and handled across DRT systems. 3.5 Rural DRT—Performance Data to Measure Transit Impact In addition to the more traditional performance data used to assess DRT service, less tradi- tional data elements may also be used particularly if there is an interest in incorporating social values in DRT performance assessment. Such data elements recognize that rural transit often serves predominately those who depend on transit and that the more traditional data elements such as passenger trips and vehicle-miles do not necessarily capture the role of rural transit in improving the quality of life for rural residents. These data elements might include, for example, the number of riders transported to congregate meal sites for a nutritious meal or the number of rural residents able to access improved health care because of rural transit. Such less traditional data relate to the impact of rural transit, and can be used for impact mea- sures that weigh the results of rural transit. The impact or result of rural transit might be measured, for example, by assessing the economic impact of ensuring access to employment for a specific number of residents who previously had received welfare. The types of impact data that a rural DRT system might want to collect depend upon what it wants to measure. If there is a local inter- est in improving transportation for lower income seniors, then data collection should target how the rural DRT system is improving seniors’ access within the community and the results of that access such as total number of trips to the senior center, total number of trips to the grocery store, and so forth. This Guidebook does not suggest any specific data elements for measuring transit impact since such data and resulting measures are more properly determined at the local level, based on the particular community issues with which the rural transit system is involved. 22 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance