ABSTRACT
Developing
new intelligent transportation systems to meet unusual challenges in rural areas
where ITS implementation is limited is difficult in part because there are few
similar examples to use as guides . In
such situations, it is critical to conduct a thorough analysis of user needs to
provide a strong foundation for the development process. This paper describes the analysis conducted
to support the user needs assessment for the Northern Shenandoah Public Mobility
Project in Virginia. This project, which
seeks to improve the coordination of human services transportation, is unique
in low-density, rural regions. The report
illustrates the importance of the use of geographic information systems (GIS)
in supporting the necessary analyses.
INTRODUCTION
Rural
areas face transportation challenges that are quite different from the challenges
faced by urban areas, but they too are looking to advanced intelligent transportation
systems (ITS) technology to help. As with any ITS implementation, success depends
largely upon properly understanding the problem and what ITS solutions might alleviate
that problem. In other words, long before
an area decides to pursue an ITS solution, it should conduct a thorough investigation
into the nature of the problem, conduct an analysis of the needs, and study a
wide range of alternative solutions. The following paper describes in detail how
a Geographic Information System (GIS) was used to analyze a rural transportation
problem in the Northern Shenandoah Valley of Virginia.
The
Northern Shenandoah Valley Public Mobility Project is an ongoing effort to create
a coordinated human service transportation system for the Northern Shenandoah
Valley region of Virginia using advanced intelligent transportation systems technology.
A sampling of the coordination elements under investigation include:
- networked computer aided
dispatching;
- ride-sharing for the clients
of the participating human service agencies;
- van-sharing between the
agencies for on-going or event specific transportation;
- flex-routing and demand
responsive transportation to maximize existing human service transportation routes.
The
Public Mobility Project is a unique ITS public transportation effort in that it
focuses on transportation provided by human service agencies. The primary mission of these agencies is to
provide a wide range of services to special needs clients, such as the elderly,
disabled, etc. Therefore, transportation
is not the primary objective of the agencies. However, they must invest substantial resources
toward providing transportation services in order to accomplish their goals, and
they recognize the need to improve efficiency. Furthermore, this project takes place in a
region that can generally be characterized as rural, although not isolated.
Socially and economically, it is tied to the Washington, D.C. metropolitan
area, which is located some 35 miles to the east.
With the exception of Frederick County, which includes the city of Winchester,
none of the counties have populations greater than 50,000.
Northwestern
Community Services and the Committee for Coordinated Human Service Transportation
have led this effort. The Smart Travel
Laboratory of the University of Virginia’s Center for Transportation Studies has
provided support by analyzing the operational feasibility of ITS alternatives.
The purpose of this paper is to describe the geographic information systems
analysis conducted by the University of Virginia. It is intended to provide guidance
and information to those considering similar projects in the future.
Please note that since this analysis was completed, the project has progressed
and the agencies are currently in the process of developing a computer-assisted
coordination support system.
GIS
ANALYSIS BACKGROUND
The
purpose of the GIS analysis and report was to make observations and recommendations
based on an analysis of the routes and unmet needs provided by the human service
agencies participating in the Northern Shenandoah Valley Public Mobility Project. The analysis attempted to answer, from an operational
standpoint, the question:
can coordinated transportation work and what kinds of coordination are
feasible? The sections that follow contain observations drawn from the GIS
analysis and a number of static maps generated from the project to illustrate
those observations. It also contains an
analysis of the various alternative products of the project.
GIS
is an information technology that allows for the integrated analysis of spatial
and attribute data. It has proven to be
valuable in varied application areas, ranging from natural resources, to community
planning, to construction site analysis. In this effort, UVA utilized one of the most widely used GIS packages
on the market, ESRI’s ArcView.
PROJECT
COMPONENTS
The
GIS project that UVA prepared in developing the report consisted of several different
components. First is the existing route
data for human service agencies. Tuesday was chosen as a representative day and each route was represented
on a map as a series of connecting line segments. Attribute data was also collected for each
route segment to store the number of empty seats on the vehicle. This data was visualized based on the following
scheme: a dark blue line represented the
largest number of empty seats and faded to light blue as those seats were filled.
Line segments with no empty seats were colored red so that they would stand
out. Each agency’s routes were created as a separate
layer so that they could be easily made visible or invisible on the map.
Second, the unmet need (spatial region) for each agency was drawn as a
set of polygons on their own map layer so that they too could be easily turned
on and off. Third, a layer was created that displayed all the routes color-coded
by client type. Finally, a three-mile
buffer was created around all the routes to demonstrate areas of the region that
no agencies have the potential to service with current routes/vehicles.
NATURE
AND QUALITY OF THE DATA
Like
any other analysis, this effort relied on the data that was used to support it.
The data for this GIS analysis was obtained through several rounds of interviews
and site visits with the participating agencies as well as mailed surveys. There was quite a bit of information available about routes that
run on a more or less regular basis. Most
agencies provided data for what they felt was a typical day for their vehicles.
Not surprisingly, there was very little information available on demand
responsive service because there was simply very little demand response service
provided. Finally, the data on unmet need was general
in nature but fit well into the context of the project.
The
heart of the GIS project was the representation of the regular routes that the
agencies run. Tuesday was chosen as a
representative day and the routes were grouped into those that occur in the morning
and those that occur in the afternoon. The
morning and afternoon classification was quite sufficient for the purpose of identifying
trends and making observations on the possibility of coordinated transportation.
GIS
ANALYSIS RESULTS
The
GIS analysis revealed a number of features of interest in the context of cooperative
transportation efforts.
·
A significant portion of the
region is accessed by the existing routes (Table 1 and Figure 1).
This can be seen with the map
buffering all the routes. The highlighted
blue region spans three miles on either side of all the routes.
This three mile buffer covers 65% of the total area of the planning district. If a two mile buffer is used, 53% of the planning district is covered
and 34% is covered by a one mile buffer.
Table 1: Portions of total planning district land area
covered by route buffers.
|
Total area of the Lord
Fairfax Planning District: 1650 square
miles. |
| | Area | Percent of Total | |
| 3 mile buffer | 1070 | 65 | |
| 2 mile buffer | 870 | 53 | |
| 1 mile buffer | 570 | 34 | |
Figure 1. Three mile buffer around all the routes.
·
One way to look at the routes
is to see them as connecting small towns (Figure 2).
For
instance, Northwest Community Services runs a good Winchester – Front Royal connector
and Northwest Workshop runs a Winchester – Berryville connector. In fact, many towns in the planning district
are connected to their neighbors in this way.
This way of looking at the routes illustrates one potential method for
coordinating service.
Figure 2. Route/Town Connectivity
·
Shenandoah Agency on Aging
and Northwest Community Services are the two largest agencies (Figure 3).
These
are the agencies that could play the most important role in a cooperative system
because they cover the most ground. The
Shenandoah Area Agency on Aging has 66% of the total route segments and Northwest
Community Services has 14%.
Figure 3. Agency Shares of Route Segments
·
There
are many empty seats on the majority of the routes (Figure 4).
A
graduated color scale was used to highlight where each agency has empty seats
on their routes. On the typical Tuesday
morning, only 12 of 262 route segments are filled (5%). Northwest Community Services, Shenandoah Area
Agency on Aging, Shen-Paco Industries, and Northwest Workshop all stop at individual
residences or group homes and pick up one or two people at a time to take to some
destination. Their vehicles are then full
only from the last stop to the destination – they have empty seats available on
every other route segment. Grafton, Warren County Workshop, and Shenandoah Valley Community
Residence transport all of their clients from a single pick up point to another
destination, but do not always fill up their vehicles. These are the empty seats that could be filled
by another client in a rideshare situation.
Figure
4. Empty Seats
ANALYSIS OF ALTERNATIVES
Each
of the alternative system approaches described in the introduction section will
be addressed below. In the broadest possible
terms, the data suggests that the availability of empty seats is not a problem.
Making it easy to find those empty seats would need to be part of the job
of a system that facilitates cooperative transportation efforts.
Secondly, if no new routes are created, there will still be areas of need
that cannot be reached.
·
Van
Sharing
Van sharing has been a very unpopular idea thus far
among the participating agencies. However,
the morning and afternoon schedule does indicate that there is likely significant
vehicle down time. From conversations
with all of the agencies, it seems that there is seldom a driver available during
that vehicle down time, though. Some agencies only hire drivers part time (Northwest
Community Services). Other agencies have
other responsibilities for their drivers during vehicle idle time.
For instance, Shen-Paco and Warren County Workshop drivers are also work
supervisors, Grafton drivers help in the school, and Shenandoah Area Agency on
Aging uses their drivers during midday down time to deliver meals to home bound
seniors. Vehicles also seem to be available during the
evening and weekend hours, though many are used for sporadic outings and chores.
·
Ride Sharing (Route Optimization)
Ride sharing might be considered “route optimization”
which could either be approached at a system wide level or on a smaller scale
to meet specific objectives. In either
case, agencies would cooperate on some of their regular route service by giving
rides to clients from another agency while that other agency gives rides to some
of their clients. Goals might be to reduce
total route time and distance while giving more people an opportunity to participate
in programs. Looking at the ArcView project, there is considerable overlap in
where routes are run. It remains a question
whether there is sufficient overlap in times and direction for a large-scale optimization
effort to be effective. However, a major optimization of routes would change dramatically
the way that agencies handle their transportation and would likely be unpopular
for that reason. On a small scale, agencies
could look for specific situations in which they could cooperate in order to reach
previously unmet need. For example, Figures
5 shows how Northwest Community Services pass through areas of need identified
by the Frederick Department of Social Services.
In fact, 67% of the areas of unmet need fall within three miles of an existing
route, indicating that at some level route optimization could be quite effective.
Figure 5. Unmet need of the Frederick Department of Social
Services and the routes of Northwest Community Services.
·
Flex
Routing and Demand Responsive Transportation
Flex
routing and demand responsive transportation are perhaps the most feasible of
the above ideas. The fact that so much
of the planning district is covered by the routes is certainly encouraging, even
if the times that vehicles travel each route are limited. Given the degree of variability inherent to
many of the routes, a demand responsive system would likely work best for clients
who have some flexibility in their trips. This specifically indicates that people who need transportation
to and from a job on a regular basis may have difficulty with the options available,
whereas those who need a ride periodically to a medical appointment or shopping
center might have more success. Agencies could provide morning trips out with
a return in the afternoon. These routes
could be viewed as connecting towns with departure and arrival times in each town.
Note, however, that the nature of many of the routes is to stay within
a fairly limited area. Though a demand
response client may be able to go between a few towns within their county, they
may not be able to go to the far side of the planning district. It would probably not be practical to attempt
to arrange transfer trips.
·
Computer
Aided Dispatching
Finally,
the dispatching issue is essentially a question of communication. Though it is not geographic in nature, some
information on current dispatching procedures did come up in the course of gathering
route data. Almost all the agencies accomplish
their dispatching through informal means. Those with regular routes only have to keep
track of daily variability, which is often best accomplished by having the driver
or someone close to the driver in the organization handle minor route changes.
For example, Shenandoah Area Agency on Aging has each of the six senior
centers do their own dispatching. Shen-Paco’s
dispatching is very informal and changes are made by calling an administrator
in the office. By a conservative count,
perhaps 18 dispatch locations are involved.
Whatever
the product of this effort is, it is sure to rely heavily on clear communication
so the effective use of communication tools will play an important role.
Given the relatively small number of participating agencies and the relative
scarcity of resources (time and money), the product should be kept fairly simple.
However, there is a significant amount of complexity to the issues at hand. An effective way to tackle that complexity
would be to develop a system that allows dispatchers to quickly view some generalized
options and then contact the other dispatchers in question. In other words, the system would provide general
information then get the right people talking to each other in order to work out
all the necessary details. This approach
would deal effectively with daily variability and other tricky factors such as
client compatibility by letting the people involved make those decisions based
on a first hand knowledge of the factors involved. The decentralized nature of dispatching would
be one of the challenges to overcome in such a system.
CONCLUSION
As
described in this paper, experience in the Northern Shenandoah Mobility Project
indicates that GIS is an effective tool to use in identifying the need for, and
requirements for, ITS solutions for public transportation challenges in rural
regions. It is recommended that a rigorous
needs assessment, using tools such as GIS, be conducted to support the development
of unique ITS systems.
