New York, NY
Sequencing Decisions in Paratransit System Design
INTRODUCTION: THE NATURE AND STATE OF THE PROBLEM
Required by the Americans with Disabilities Act for nearly a decade, and with almost three decades of similar experience preceding it, complementary paratransit services should “have the bugs worked out.” In contrast, their problems seem to be growing:
- Not only has efficiency not risen remotely to the levels of many pre-ADA systems, it is either stagnant or declining.
- System reliability remains a problem, despite the use of GPS-based technologies to create the schedules.
- Complaints have swollen to a point where some Congresspersons and city council members maintain a “collection” of them.
- Transit agencies and their contractors are routinely “toasted” by local press and media.
- Accidents are increasing sharply, and damage awards growing exponentially.
- Many insurance underwriters are refusing to even “write paratransit business.”
- A number of systems are mired in Class Action suits – and in some cases, continue to violate provisions of the ADA while operating under Consent Decrees they had agreed to in the settlement of those suits.
Missing from the paratransit landscape almost entirely are many of the clever service concepts pioneered in the Late 1970s and Early 1980’s, mostly in suburban and rural service areas. While the dedicated vehicle portions of these systems typically provided four or five passenger trips per hour, often with 98 to 99 percent on-time performance, some systems provided more than 10 trips per hour. In contrast, many of Today’s complementary paratransit systems provide less than one – despite an army of lead agency management personnel and an arsenal of advanced technology.
THE NATURE AND PACE OF EVOLUTION
One argument often made in defense of paratransit’s problems is that ADA requirements are unreasonable, and that such rigorous standards cannot realistically be met at all, much less at reasonable cost. Considerable time is spent griping about the ADA being an “unfunded mandate.” Unfortunately, these excuses do not account for the fact that many pre-ADA systems applied similar standards to the transportation of a far broader base of clients – and transported them far more efficiently. One can only conclude from this that many lessons of these earlier systems were never conveyed or absorbed.
Many laws, institutions and organizations share the blame for these problems. However, the ADA is not one of them: Some form of paratransit service existed in practically every transit service area long before the ADA was promulgated. However, one important distinction between pre- and post-ADA paratransit systems helps to explain the problems so many current systems have:
- With their roots in “Sixties” programs like subscription service for the Elderly and feeder-to-rail service (e.g., the Haddonfield demonstration project), pre-ADA paratransit systems evolved over a long period of time from essentially a “clean state” with few technical requirements. The primary concern of the more than 60 different Federal programs providing funds for paratransit service by the mid-1970’s was eligibility. Among the few technical constraints these programs contained, some were so ridiculous that system officials chose to ignore them rather than punish their clients by complying. (For example, systems receiving Area Agency on Aging funds from Title III had to provide a minimum of 2.0 trips per hour – a benchmark applicable from Manhattan to rural Alaska.)
- Post-ADA paratransit systems, in contrast, were mandated as a matter of law, with specific constraints defining the parameters of key service quality issues. Transit agencies were given a generous, five-year timetable for full implementation. But when the systems began operating, they had to operate immediately in accordance with most of the requirements.
The importance of the “developmental environment” enjoyed by pre-ADA paratransit systems cannot be overstated: Free from detailed institutional mandates and threats (e.g., losing one’s transit funds, exposure to Class Action suits), pre-ADA paratransit officials had the opportunity to continuously try new approaches, and assemble, reassemble and refine their systems time after time, function by function. As the principles which governed performance began to emerge, the best systems were slowly reshaped to reflect them. In the most enlightened cases, system decision-makers seized the opportunity to prioritize those decisions which most affected performance, and to sequence them to optimize the system’s overall design. Over time, system performance improved not only because all the “pieces” were present and made sense on their own terms, but because they fit together.
Figure 1: Paratransit System Decision-Making: Interrelationships and Sequencing presents the model this author developed for the USDOT in 1980.(1) The complete model – which included a volume of text describing the interrelationships among each set of intersecting variables – accomplished four distinct objectives:
- It identified the key decision-making areas within a paratransit system’s structure, processes and functions.
- It prioritized the order in which decisions in each area should be made, so that the most important, often difficult-to-change decisions drove decisions either less important or more malleable.
- It identified every decision-making area that affected other decision-making areas and, conversely, each area affected by decisions in other areas.
- It identified the specific factors involved in each decision-making area, and described how they affected decisions in other areas.
This model did not require one to scrap the system. Rather, it provided the key to redefining the variables so that they could be fitted together. To understand how such a model works, it is helpful to see what happens in common practice without such a model.
MISMATCHING PARTS AND THEIR IMPACTS
Figure II presents a schedule for two shared-ride pickups and drop-offs assigned to a simple minivan containing two wheelchair positions configured in tandem. The task seems simple enough: Two passengers, and a vehicle large enough to carry them. Certainly no scheduling software program is going to “flag” this situation. So, what’s the problem?
The problem is that this simple schedule cannot be accommodated by this simple vehicle. The pieces “don’t fit together.” Look what happens when a driver is given this schedule and vehicle:
- If the driver secures Passenger A to the front tie-down position, when he gets to Passenger B, he will not be able to load him, much less secure him to the rear tie-down position, since Passenger A is in the way. To load Passenger B thus involves a three-step sequence:
- Untie Passenger A and wheel him to the rear tie-down position
- Re-secure Passenger A to the rear tie-down position
- Load Passenger B and secure him to the front position
- If the driver secures Passenger A to the rear tie-down position, he will be able to load Passenger B directly (avoiding the steps above) and secure him to the front tie-down position. However, when the vehicle arrives at Passenger A’s drop-off, the driver will not be able to unload him because Passenger B is secured to the front tie-down position, blocking Passenger A’s exit. So the driver must now perform three loading/unloading operations:
- Untie and unload Passenger B from the front position, move him out of the way, and set his brakes
- Untie and unload Passenger A
- Reload and re-secure Passenger B (to either position)
Even without any other complications, this scenario is an accident – and a serious law suit – waiting to happen. Even if the driver solved this “block-set” at the first drop-off so that he did not experience two three-step aberrations – the driver would be significantly behind schedule and operating in the “Land of Murphy’s Law.” Efforts to capture the lost time might include:
- Rapid acceleration, deceleration and/or turning
- Unnecessary merging and weaving
- Incomplete wheelchair securement
- Improper wheelchair adjustment and tightening
- Improper passenger securement to the wheelchair
- Incomplete passenger assistance (e.g., not physically escorting passengers to and from their origins or destinations)
- Failure to record information on the log
The ease with which this simple situation can unravel helps to explain the litany of errors and omissions one generally finds in a wheelchair-related accident. Keep in mind that most wheelchairs would remain in place without being tied down at all if the driver (a) set the wheelchair’s brakes and (b) drove carefully and reasonably. And most wheelchairs can be securely fastened by properly fastening and adjusting restraints to only three of the four wheel positions. (Otherwise, how could a motorized three-wheeled scooter be secured?) So when a wheelchair comes loose from its moorings, and its occupant flies about within the passenger compartment, a serious number of errors and omissions almost certainly occurred.
The reality is, one cannot give a paratransit driver a puzzle and expect him to safeguard the welfare of his passengers. Unless the schedule allows enough time for three loading/unloading sequences at all but the first pickup and last drop-off, the combination of schedule and vehicle cited in the example above is a disaster waiting to happen. From the perspective of a personal injury attorney and his or her expert, combining these two elements is already a critical error – an error made considerably high up within the management hierarchy. That more errors and omissions will occur as a result of it is, in legal parlance, “reasonably foreseeable.”
Comparing this accident scenario to the Model (see Figure 1), one can identify the many improperly-coordinated variables which contributed to it:
- Eligibility I
- Roles & Responsibilities
- Eligibility II
- Organizational Responsibilities
- Scheduling and Dispatching
Changes in any one of these areas could have helped avoid the dilemma above:
- Knowing wheelchair occupants would be transported, system designers could have “thought through” its implications in subsequent decision-making areas (Eligibility I).
- Schedulers, dispatchers, trip reservation personnel and supervisors (including those assigning vehicles to “runs”) could have made sure the vehicle deployed had enough wheelchair positions to accommodate the schedule (Roles & Responsibilities).
- Knowing the percentage of wheelchair occupants among total clients, system designers could have coordinated scheduling, dispatching and trip reservation procedures with the preparation of vehicle specifications (size, wheelchair capacity, etc.) and the purchase of the proper mix of lift- and non-lift vehicles in the fleet as a whole (Eligibility II).
- Supervisory and monitoring personnel could have been assigned to identify discrepancies between schedules and vehicle assignments. Or aides or attendants could have been assigned to runs unavoidably involving intermittent “off-loading” and re-loading (Organizational Responsibilities).
- Given the number of wheelchair occupants and the system’s pre-scheduling format (which would affect the likelihood of overlapping wheelchair occupants), larger vehicles or those with more securement positions could have been ordered, or their presence could have comprised a larger proportion of the overall fleet (Fleet).
- Securement positions could have been located side-by-side rather than in tandem. Or additional securement positions could have been located beneath fold-down seats (Equipment).
- Schedulers and dispatchers could have been trained to spot pending problems and, when “problem runs” had already begun, they could have “TX’d” other vehicles to transport specific subsets of their passengers (Scheduling and Dispatching).
- Drivers could have been trained to expect occasional anomalies, to spot them in advance, to radio in for assistance, or to at least foresee the “traps” poor system design, vehicle assignment, scheduling and other ineptly-performed functions might generate: A trained driver might have at least avoided the three-step loading sequence at the second pickup in the example cited above. (Training).
The frightening thing about this example is that, as complex as it turns out to really be, real-life operations are several orders of magnitude more complex. Typically (if not always):
- Schedules are generated by software programs incapable of the subtleties essential to the execution of service.
- Schedules – computerized or otherwise – cannot specify the appropriate seating position for each passenger.
- Schedules are devised by a different organization (commonly the lead agency, a software development firm or a broker) than the one responsible for executing it.
- The scheduling, dispatching and trip reservation process is fragmented, with no individual (much less a live Earthling) ever performing more than one of them for a single passenger.
- Poor schedule adherence is penalized by “liquidated damages” – a device which forces operating agency management and personnel to bastardize the performance of other system elements to minimize them.
- Drivers and management personnel are pre-occupied with administrivia and besieged by useless or redundant technology (e.g., mobile data terminals).
- Log review is either not performed or not performed regularly.
- Back-up drivers or posted vehicles are not poised for duty – generally because the lead agency doesn’t recognize their necessity or is unwilling to pay operating agencies for them.
- Operating agency personnel are besieged by calls regarding late pickups.
- Lead agency personnel are besieged with complaints and public relations catastrophes.
When a paratransit passenger is injured or killed, his personal injury attorney (or that attorney’s expert consultant) often finds all of the above – and sometimes scores of errors and omissions related to the injury or death scattered throughout the management hierarchies of both the lead and operating agencies.
Even apart from the ramifications a simple discrepancy like that illustrated above can have on passenger safety, the mismatching of system elements that typically occur are far more severe, far more complex and often more subtle, to begin with:
- A lead agency encourages clients to pre-schedule their trips – then subcontracts out service to a taxi company. (Advanced reservations are anathema to taxi dispatchers who don’t schedule any trips but, rather, save them as “time calls” and, at the last minute, sort (or “dispatch”) them into the nearest available vehicles.)
- Through coordination, a system expands eligibility to non-disabled elderly residents – most of whose trips can be provided on a subscription basis – then limits “standing orders” as a means of restricting ridership.
- A system purchases a fleet of expensive, fuel-guzzling, lift-equipped minibuses, then provides most of its trips on an immediate-response basis –and compounds this waste by setting unreasonably low maximum ride time limits which make trip-sharing a near impossibility.
- Another system – like the one in our example – orders smaller ramp-equipped minibuses for a largely subscription-based service top-heavy with wheelchair occupants, fails to grasp this relationship in developing scheduling algorhythms, and as a result, drivers have to constantly remove wheelchair occupants at fellow passengers’ stops because the chairs opposite the door block the entry and exit of other chairs.
Many costly errors like these have been institutionalized by charlatans masquerading as consultants, as well as by lead and operating personnel introduced to paratransit largely through scheduling and dispatching software. Many self-proclaimed paratransit “experts” actually claim that immediate-response service is more efficient than subscription service – when both empirical evidence and common sense demonstrate the opposite.
The problems such conditions create is exacerbated where operating agency personnel possess genuine knowledge but don’t have the authority to effect the changes their system needs to address their problems. RFPs permitting bidders to actually design the systems are rare. When they exist, “seasoned” bidders don’t risk proposing models different from the pre-conceived notions of lead agency personnel. Once operations unfold, and problems begin, this conflict often escalates into a war over liquidated damages and their “validation.” All too often, “violations” reflect inept or sloppy decision-making and fatally-flawed system design.
If the pieces of a paratransit service don’t make sense on their own terms, they cannot fit together into a coherent system. Yet even if the pieces do make sense on their own, the service will still have serious and multiple problems if the pieces don’t fit together into a coherent system which makes sense, and can be articulated, at the operating level.
Where the pieces do not fit together, accidents will follow. Where accidents exist, law suits will follow. The longer system officials fail to address these fundamental problems, the worse their services will become, and the more they will cost directly (from poor efficiency and higher insurance premiums) and indirectly (from law suits and damage awards). As transit agencies are continuing to discover, because system complexity increases exponentially with ridership, so too does the cost of failure.
Acknowledgements and Disclaimers
The views expressed in this paper are exclusively those of the author, and do not necessarily reflect those of the American Public Transportation Association or any of its other members.
- Einstein, Ned. B., Special Paratransit Services for Elderly and Handicapped Persons: Volume 3: Decision Manual for System Design. Washington, D.C.: Office of Planning Assistance, Urban Mass Transportation Administration, 1981.