An Exploratory Study of Encumbered Passengers on Fixed Route Buses

Abstract

Although the presence of disabled passengers on fixed-route transit affects operations, their data are often ignored, aggregated into a broad “encumbered” category, or marked as outliers in analyses of service performance measures. This study examines boarding and alighting characteristics of passengers encumbered with mobility aids (wheelchairs, scooters, walkers, and canes) and accessories (bicycles, strollers, and trailers) on fixed route buses and compares them to ambulatory unencumbered passengers. Data from detailed task analysis of video recordings of boarding and alighting activity at 1,327 bus stops from a sample of 29 bus trips on two fixed routes in Ann Arbor were analyzed. Results indicate that encumbered passengers did not travel in the morning peak period, but with the exception of wheelchair users, did not avoid the p.m. peak period. Almost all passengers with mobility aids, about half of those with accessories, and about a third of ambulatory passengers exited through front doors. The overall average dwell time was 19 seconds and 15 seconds when no encumbered individuals were boarding and alighting, respectively. Average dwell times ranged from 28 to 137 seconds when encumbered individuals were among those getting on or off. Examination of dwell times by specific mobility aid and accessory suggest that grouping passengers into: ambulatory with no accessories, encumbered with no ramp needed, and ramp needed might be a simple practical approach to including the effects of the broad diversity of transit passengers, both with and without disabilities into dwell time analyses, and in turn into measures of transit operations.

INTRODUCTION

Recent decades have seen increased opportunities for individuals with disabilities to be more active socially in employment, education, and recreation. Much of this progress can be attributed to the passage and enforcement of federal accessibility legislation such as the Americans with Disabilities Act (ADA) (Title II, Part B) (1) that require public transportation agencies to make fixed-route services accessible and usable by persons with disabilities, and to provide complementary paratransit service to those individuals who cannot use the fixed-route services and meet specific regulatory criteria.

Since the passage of the ADA in 1990, transit systems have been striving to accommodate individuals with disabilities. By 2010, 98 % of the fixed route buses in the US were accessible (2). At the same the demand for paratransit services increased from about 15 million in 1991 to 67 million in 2008 (2). The costs to a transportation agency of providing trips by paratransit are significantly higher than providing trips by fixed-route service. The Government Accountability Office (GAO) reports that in 2010 on average a paratransit passenger trip cost $29.30, whereas a fixed-route passenger trip cost $8.15 on average (3). However, despite increased accessibility of the US fixed-route transit fleet, demand for paratransit service is growing (4) and transit agencies are looking at ways to attain a reasonable balance in providing fixed-route and paratransit services to people with disabilities.

To this end, strategies that enable and promote the use of fixed route transit by people with disabilities were developed and published by the TCRP in Report 163 (4). The TCRP report recognized that “some people with disabilities cannot use fixed-route transit even if it is fully accessible and even if the environment around stops and stations is accessible” (pg. 2) but notes that there are people with disabilities who may have the ability to use fixed-route transit services, for at least some of their trips, who do not use the fixed-route systems. Several reasons for this are given, not the least of which are, limited or lack of experience traveling by bus and routes may not go to desired destinations (4). However, also listed among reasons for the nonuse of fixed-route services are the attention and disruption to the service perceived by disabled riders that may result from delays they cause. In a study of wheelchair tie-down comfort and safety, Wrestrand (5) notes that wheelchair users might see themselves as being a burden to other passengers.

Most ambulatory passengers are considerate of the additional time taken by riders with disabilities in boarding and alighting, but there are some who are not as patient or understanding (e.g. 6). In a national survey of almost 2,000 disabled fixed-route riders conducted as part of the TCRP research (4) 22% of respondents indicated that attitudes of other passengers were very important to them in their decision to use fixed-route transit.

But what is the effect of disabled riders on fixed-route transit operation? Is it different than that of passengers with accessories and does it vary by the type of mobility aid the rider uses? Dwell time, defined as the time a bus is stopped to load and unload passengers, is accepted as instrumental to transit operations (7). It depends on the number of individuals getting on and off, and their speed in doing so. Models to predict expected dwell times are developed from empirical data, which usually contain few cases of passengers with disabilities or special needs. Early dwell time models included only the number of passengers boarding and alighting (e.g., 8–10). Fare structures were later included (e.g., 11, 12), as was time of day (e.g., 13) and passenger load and door crowding (e.g., 14–17). However, because individual with disabilities are considered a very small portion of passengers on fixed-route buses, their data are often treated as a rare event or an outlier and excluded altogether from the dwell-time analysis (18).

Few dwell time models account for passengers with disabilities in any way. Lift operations were included in models by Dueker et al. (19) and Rajbhandari et al. (20). However, these models do not account for disabled riders who do not need a lift but still need more time than fully ambulatory passengers to board and alight. Other approaches to dwell-time modeling combine disabled passengers together with ambulatory passengers with accessories such as bicycles, strollers, trailers and classify all as encumbered (e.g.,21). Still another approach is to treat the presence of disabled or other encumbered passengers as unexpected or atypical events (e.g., 7).

Consequently, the influence of passengers with disabilities is either totally or partially ignored or aggregated into a broad category. There is a need to understand more about the use of fixed route service by disabled individuals using a variety of mobility aids. It is especially important to know how such riders compare with ambulatory riders, and with those with additional accessories in the measures associated with dwell time. Such knowledge can help transit operators make their service more efficient for all passengers and encourage more use of fixed transit among individuals with disabilities who are able to use the system.

Objectives of this paper

The objective of this paper is to further the understanding of how people with mobility aids use fixed-route service and to explore a practical way of including the effects of passengers with disabilities into dwell time analyses and other measures of service performance.

DATA AND METHODS

This study is part of a larger research effort concerned with the use of public transportation by people with mobility impairments in Ann Arbor, Michigan. The Ann Arbor Transportation Authority, a.k.a. TheRide, provides fixed route and paratransit services in the cities of Ann Arbor and Ypsilanti and nearby townships. All of TheRide’s fixed-route buses are low-floor, can kneel to the same height as the curb, are equipped with ramps at the front door that deploy for barrier-free access, and can accommodate persons who use mobility aids such as wheelchairs, scooters, and walkers. Drivers assist wheelchair users in boarding, alighting, and securing wheelchairs in one of the two designated wheelchair locations. Passengers eligible for TheRide’s paratransit service ride the fixed-route bus for free, and reduced fares are available to individuals with other mobility impairments. The portion of persons with disabilities in the total fixed-route ridership of TheRide has been estimated as 2.5%, and the ratio of fixed-route ridership by persons with disabilities to paratransit ridership was 1.12 in 2011 (4).

The buses are also configured to accommodate passengers with accessories such as strollers and bicycles. The transit agency created a designated area on-board for strollers, which can also be used by passengers with walkers, by taking out one of the curb-side seats on all of its buses. Buses are equipped with two bike racks available without additional charge or special permit required. No driver assistance is provided for securing and unloading bicycles.

Data Collection

As part of the larger study, detailed data about the boarding and alighting of fixed-route bus passengers were collected through a systematic task analysis of on-board surveillance video records from a sample bus trips on two service routes, known to serve passengers with mobility aids. The study was approved by the University’s institutional review board. Figure 1 shows the routes maps of the two routes in the study, route 4 and route 6. Both routes travel between the Ann Arbor and Ypsilanti Transit Centers and run through densely developed areas past medical facilities, hospitals, major shopping areas, community centers, and university locations. Frequency of service on route 4 is from 5 to 10 minutes during peak periods, between 10–20 minutes during morning and afternoon off-peak periods and 30 minutes in the evening. Frequency of service on route 6 is 15–30 minutes during the day, and 1 hour in the evening.

Figure 1.

Service Map Showing TheRide’s Routes 4 and 6.

TheRide staff provided the research team with on-board surveillance videos of 96 bus trips from route 4 randomly chosen from 4 week-days of operation and 23 bus trips from route 6 randomly chosen from 3 week-days of operation. During those days, TheRide operated 162 bus trips/day on route 4 and 57 bus trips/day on route 6. Thus, the 119 bus trips in the sample drawn for the systematic task analysis account for 14% of bus trips on routes 4 and 6 routes during the study days. Because the present study is concerned with passengers with mobility aids and accessories, only bus trips that contained at least one stop at which a passenger encumbered with a mobility aid or accessory boarded or alighted were retained. Of the 119 sampled bus trips, 29 trips (24.4%) satisfied this criterion.

Data for the present study were developed from the variables collected in the task analysis, and contain information for each stop in the 29 bus trips at which passengers boarded or alighted. The data for each bus stop include: stop identification and location, date, time of day, start of dwell time, end of dwell time, the number of passengers boarding and alighting by their mobility status (ambulatory, wheelchair, cane, walker, scooter), number of ambulatory passengers with an additional accessory (bicycle, trailer, stroller) or carrying an infant, method of fare payment, doors used when alighting, and any unusual incidents.

We first examined the incidence of encumbered passengers among all passengers in our sample of bus trips by their mobility aids and accessories and examined variables that may contribute to dwell time such as the times of day of travel, methods of fare payment, and use of the front door when exiting the bus. Next, we examined the bus stops classified by the make-up of the passengers boarding or alighting. The stops were classified as, stops with only ambulatory passengers, stops with at least one encumbered passenger, stops with only ambulatory passengers with at least one passenger encumbered with an accessory, and stops at which at least one passenger boarding or alighting used a mobility aid. The average number of passengers boarding and alighting, average passenger load (i.e., the number of passengers on-board), and the distributions of dwell times were obtained and compared. These were further examined by the type of mobility aid or accessory used by passengers.

RESULTS

There were a total of 1,327 bus stop events on these 29 bus strips, with 119 of them at the Ann Arbor or Ypsilanti terminal and 1,208 at bus stops along the routes. In the rest of this document the bus stops on route are referred to bus stops, and the stops at the terminal are referred to as terminal stops.

Table 1 shows the number of passengers boarding and alighting at the terminal stops and bus stops by their encumbrance category. Overall, there were 2,615 passengers boarding and 2,569 passengers alighting. The numbers are not equal because it is possible to stay on the bus at the terminal, and some of the buses go to other routes upon completion of a bus trip on route 4 or 6. Of the total passengers boarding, 95% were ambulatory with no accessory, 2% were ambulatory but had an accessory or infant in arms, and 3% had a mobility aid. Thus, about 5% of the passengers can be considered encumbered in some way.

TABLE 1.

Number of Passengers Boarding and Alighting at Terminal Stops and Bus Stops

Passenger categories	Terminal stops		Bus stops		Total
	Boarding	Alighting	Boarding	Alighting	Boarding	Alighting
All	1,129	903	1,486	1,666	2,615	2,569
Ambulatory no accessory	1,075	850	1,418	1,612	2,493	2,462
Ambulatory with bicycle	9	11	8	7	17	18
Ambulatory with stroller	1	2	2	2	3	4
Ambulatory with infant	0	2	1	0	1	2
Ambulatory with trailer	7	10	19	13	26	23
With mobility aid: Wheelchair	8	7	8	8	16	15
With mobility aid: Cane	13	9	20	12	33	21
With mobility aid: Walker	13	9	8	11	21	20
With Mobility aid: Scooter	3	1	1	1	4	2

Among passengers boarding with mobility aids, 45% used canes, 28% used walkers, 21% used a wheelchair, and 5% used scooters. Trailers and bicycles accounted for 55% and 36% of accessories respectively. These proportions are not surprising. People often carry their grocery shopping purchases in these trailers, and bicycle use is greatly encouraged in the city of Ann Arbor.

Peak periods of traffic and transit demand in Ann Arbor are between 0600 and 0900 hr. and between 1500 and 1800 hr. Off-peak periods occur between 0900 and 1500 hrs. and after 1800 hrs. Table 2 shows the number of passengers boarding and alighting classified by encumbrances by time of day.

TABLE 2.

Passenger Boarding and Alighting During Peak and Off-peak Periods

Passenger categories	Total number boarding or alighting	AM Peak 0600–0900 hr.	Morning Off-Peak 0900–1200 hr.	Afternoon Off-Peak 1200–1500 hr.	PM Peak 1500–1800 hr.	Evening 1800–2400 hr.
All	5,183	310	1,046	1,380	1,352	1,095
Ambulatory no accessory	4,957	301	1,009	1,316	1,286	1,045
Ambulatory with bicycle	35	2	7	10	10	6
Ambulatory with stroller	7	0	0	1	6	0
Ambulatory with infant	3	0	0	0	3	0
Ambulatory with trailer	49	0	7	11	20	11
With mobility aid: Wheelchair	31	3	11	11	4	2
With mobility aid: Cane	54	4	3	23	11	13
With mobility aid: Walker	41	0	9	5	12	15
With Mobility aid: Scooter	6	0	0	3	0	3

Passengers with mobility aids and accessories do not appear to travel much in the morning peak period. Wheelchair users were most likely to travel during the morning and afternoon off-peak period between 0900 and 1500 hours. Individuals with accessories and other mobility aids were more likely to travel in the afternoon off-peak, peak, and evening.

The method of fare payment used by passengers affects dwell time. Table 3 shows the distribution of fare payment methods by boarding passengers.

TABLE 3.

Fare Payment Methods by Boarding Passengers

Passengers	Number of passengers	Card	Coin	Card and Coin	Did not Pay
All	2,615	63.0 %	20.6%	<1%	15.9%
Ambulatory- no accessory	2,506	64.8%	21.2%	<1%	14.7%
Ambulatory with accessory	34	58.8%	20.6%	0%	23.5%
With mobility aid: Wheelchair	16	0%	0%	0%	100.0%
With mobility aid: Cane	33	33.3%	30.3%	0%	36.4%
With mobility aid: Walker	21	23.8%	0%	0%	76.2%
With Mobility aid: Scooter	4	0%	50.0%	0%	50.0%

Clear differences were observed in the fare payment method of ambulatory passengers and those with mobility aids. Individuals qualified to use TheRide’s paratransit services are permitted to ride the fixed-route bus for free. The high proportions of riders with mobility aids who did not pay reflects this policy and indicates that this step in their boarding does not increase delay.

Use of the front door by alighting passengers increases dwell time when passengers are waiting to board through the same door. A much larger proportion of passengers with mobility aids and accessories used the front door to exit the bus than did ambulatory passengers (Table 4). Among alighting ambulatory passengers without accessories, 30% used the front door, while 51% of those with accessories used the front door. Among all individuals with mobility aids, 90% used the front door to exit the bus, and among those who did not use a ramp, 87% used the front door. Further examination found that 44% of alighting ambulatory passengers used the front door when no one was waiting to board, and 15% used the front door when passengers were waiting to board. For ambulatory passengers with accessories and persons with mobility aids, the proportion using the front door to disembark remained relatively constant whether passengers were waiting to board or not.

TABLE 4.

Front Door Use by Alighting Passengers

Passengers	Number of alighting passengers	Number (%) using front door
All	2,569	817 (31.8%)
Ambulatory – no accessories	2,466	738 (29.9 %)
Ambulatory with bicycle	18	13 (72.2%)
Ambulatory with stroller	4	2 (50%)
Ambulatory with trailer	23	12 (52.2%)
Ambulatory with Infant	2	2 (100%)
With mobility aid: Wheelchair	15	15 (100%)
With mobility aid: Cane	21	15 (71.4%)
With mobility aid: Walker	20	20 (100%)
With Mobility aid: Scooter	2	2 (100%)

We next examined the bus stops by the encumbrances of passengers boarding and alighting. Table 5 shows the number of bus stops in each category, the average number and standard deviation (SD) of passengers boarding and alighting, and passenger load.

TABLE 5.

Average Boarding, Alighting, and Passenger Loads at Bus Stops

Stops	Number of stops (% of All)	Average no. of passengers boarding/stop (SD)	Average no. of passengers alighting/stop (SD)	Average passenger load (SD)
All	1,208 (100.0)	1.2 (2.10)	1.4 (2.1)	14.5 (7.4)
Stops with unencumbered passengers	1,096 (90.7)	1.1 (1.8)	1.3 (1.8)	14.5 (7.4)
Stops with at least one encumbered passenger	112 (9.3)	2.4 (4.4)	2.2 (3.9)	14.6 (7.2)
Stops with at least one passenger encumbered with accessory	44 (3.9)	1.2 (1.8)	1.0 (1.8)	15.0 (7.1)
Stops with at least one passenger encumbered with mobility aid	68 (5.6)	2.8 (4.4)	2.7 (4.8)	15.4 (8.0)

At most (91%) bus stops along the routes only ambulatory passengers, unencumbered by accessories boarded and alighted. Persons with mobility aids and accessories boarded or alighted at 6% and 4% of bus stops respectively. Overall, there was more boarding and alighting activity at bus stops where a person with mobility aid was present. Bus stops with only ambulatory passengers, with or without an accessory averaged about 1 passenger boarding and 1 passenger alighting. However, bus stops with a person with a mobility aid averaged almost 3 persons boarding and 3 passengers alighting. The average passenger load at the bus stops was consistently about 15 people, and this did not vary by the level of encumbrance of passengers boarding and alighting.

Figure 2 shows box-plots of dwell times at all bus stops, if any encumbered passengers are boarding or alighting, and if so, whether they are encumbered with a mobility aid or accessory. The average dwell time for all stops was 19 sec. (SD=33) and a range of 2.5–619 sec. Dwell times at stops with only unencumbered passengers averaged 15 sec. (SD=22), and 53 sec. (SD=76) if there was an encumbered passenger at the stop. If the encumbrance was a mobility aid, the average dwell time was 69 sec. (SD=93) and 29 sec. (SD=19) if the encumbrance was an accessory.

Figure 2.

Dwell Time by Category of Stop Based on Passenger Encumbrance.

The activity at the stop with maximum dwell time was examined for any unusual circumstances. The bus was not ahead of schedule and there were 24 passengers on board when the bus arrived at the stop. Three passengers boarded, two ambulatory and one with a walker for whom the ramp was deployed. Although ten passengers got off the bus, all used the rear door, and it appears that the boarding of the person with the walker was the reason for the long dwell time in this case.

The large spread of dwell time at stops with mobility aid use led us look at these stops in more detail (Figure 3). The longest dwell times and largest standard deviations were at bus stops where at least one passenger getting on or off was in a wheelchair (average=137sec., SD=69) or used a walker (average=86 sec., SD=147). All the wheelchair users required a ramp, but while most of the passengers with walkers used the ramp, not all did. Average dwell time at bus stops with walker-users with ramp deployment was 106 sec (SD=167) and 26 sec (SD=7) without ramp deployment. The dwell time at bus stops with walker users and no ramp deployment was close to that of stops with cane users, almost all of whom did not use the ramp (average=33 sec., SD=45), passengers with bicycles (28 seconds, SD=15), with trailers (40 sec, SD=48), and with other accessories (38 sec, SD=23). Dwell time at bus stops with encumbered passengers getting on and off with ramp use was 113 sec (SD=117) and without ramp use was 30 sec (SD=32). This suggests that classifying bus stops by whether or not encumbered passengers are present, and if present by the use of the ramp as shown in Figure 4 might be a practical approach to form mutually exclusive and totally exhaustive groups with different effects on dwell time and other measures in operational analyses.

Figure 3.

Dwell time at Stops with Encumbered Passengers by Type of Encumbrance.

Figure 4.

Dwell Times at Bus Stops Classified by Passenger Encumbrance and Ramp Use.

SUMMARY AND DISCUSSION

This study examined the boarding and alighting characteristics of passengers with a set of mobility aids on a sample of bus trips on two fixed-routes in Ann Arbor, Michigan. Because disabled passengers on fixed-route transit are sometimes aggregated with passengers with accessories such as bicycles, strollers, and trailers into one broad encumbered category when their effects on transit operations are considered, we also included passengers with accessories in our exploration. The frequency of finding encumbered passengers on the fixed-route buses was quite low. Of the sample of bus trips available to us, only one-quarter had at least one stop at which an encumbered passenger boarded or alighted the transit vehicle. These bus trips were the subject of our analyses. In these bus trips almost 5% of the boarding passengers were encumbered with a mobility aid or accessory. However, 2.8% of the passengers boarding the buses had some type of mobility aid, which comes close to TheRide’s estimate of 2.5% disabled among its fixed-route ridership (4).

We found that in general, encumbered passengers did not travel during the morning peak period. Trips by passengers in wheelchairs were made mostly during the morning and afternoon off-peak periods between 0900 and 1500 hours, while trips by other mobility aid users and passengers with accessories were spread out in the afternoon peak and off-peak periods and evening. Thus, it cannot be argued that disabled passengers avoid the peak periods when passenger loads are highest.

Fare payment is another step that can increase time in the boarding process. The patterns of fare payment by disabled passengers are different than those of ambulatory passengers. However a large portion of the passengers with mobility aids did not pay and thus fare payment by passengers with disabilities may not contribute to delay in the boarding process.

Passengers alighting through the front door can increase the dwell time at a stop especially when people waiting to board. Overall, 15% of ambulatory passengers with no accessories used the front door to exit when passengers were waiting to board. Obviously those disabled passengers who use the ramp have to use the front door, but the other passengers with other mobility aids also favored the front door for exiting. In all, almost 90% of passengers with mobility aids used the front to exit the bus. About 50% of passengers with accessories left through the front door and this rate did not depend on whether or not passengers were waiting to board. Thus encumbered passenger use of the front door to exit contributes to longer dwell times. Encouraging use of the rear door for alighting through alternate interior configurations and usability improvements could lead to reduced dwell times (22).

The average dwell time when passengers with canes or walkers not requiring a ramp was close to the average dwell time of stops with passengers with accessories. However, average dwell times were much longer with greater variability when a ramp was needed. This finding suggests that grouping passengers into three groups, ambulatory with no accessories, encumbered but no ramp needed, and ramp needed might be a simple but practical approach to including the effects of the broad diversity of transit passengers, both with and without disabilities into the dwell time analyses, and in turn into the measures of transit operations.

It is worth noting that in the US, low-floor buses with folding access ramps are increasingly preferred over platform lifts on transit buses because ramps have lower maintenance costs and do not require the bus driver to leave his/her seat to deploy and safely operate the ramp resulting in substantially decreased dwell times. Use of platform lifts also draws unwanted attention to users making these less desirable compared to ramps. However, most low-floor buses only have access ramps installed at the front door. Further, consistently achieving a ramp slope that can be used safely and independently people with disabilities (e.g., 23) can be challenging in practice (24). The advent of automated driverless transit vehicles will only demand a higher standard of usability to insure independent use without driver assistance by passengers with disabilities.

Methodological Limitations

The analysis used data from two routes in one transit agency, and the sample size of individuals with mobility aids and accessories was small. Some of the behaviors, particularly with respect to fare payments, use of door for alighting and dwell time might differ in other locations with different fare structures, vehicle interior configurations (22), and policies.

All comparisons were based on simple summary statistics because the sample size of stops with encumbered passengers was too small to “slice and dice”. Thus, the results are not generalizable without performing similar analyses at other transit agencies, with larger samples, and more thorough statistical analyses. Nevertheless, this study demonstrates the benefits of manual methods for studying passenger activity over automated methods that rely on AVL, APC and electronic fare payment methods alone, particularly for studying user groups and travel behaviors that are extremely difficult to sample.

CONCLUSIONS AND RECOMMENDATIONS

Few dwell time models account for passengers with disabilities or other accessories in any systematic manner. This study examined the boarding and alighting characteristics of passengers with a set of mobility aids on a sample of bus trips of two fixed-routes in Ann Arbor, Michigan. The frequency of finding encumbered passengers on fixed-route buses is presently quite low, yet there are distinct difference in the durations for boarding and alighting by passengers with disabilities and those with passengers with accessories such as bicycles, strollers, and trailers. This highlights the limitations of grouping these diverse groups into one broad encumbered category.

At the most basic level, this study recommends grouping passengers into three groups, ambulatory with no accessories, encumbered but no ramp needed, and ramp needed as a simple but practical approach to including the effects of disabled passengers into the dwell time analyses, and in turn into more inclusive measures of transit operations. Incorporating knowledge of the ridership composition on different routes or specific stops along a route that potentially serves a high proportion of older adults, persons with disabilities or with accessories (e.g., hospital, independent living center, and grocery stores) could yield more accurate and reliable schedules. This knowledge can help transit operators make their service more efficient for all passengers and encourage more use of fixed transit among disabled individuals who are able to use the system.