DIAL-A-BUS (DEMAND-ACTIVATED BUS SYSTEM

A major failing of public urban transportation today is its
inability to provide adequate and attractive collection and
distribution services in lower density areas of a metropolis. In
some parts of urban areas and in many small cities and towns,
the travel demand is too small to support any transit service at
all. It is simply economically infeasible to route and schedule
present transit vehicles efficiently when only a few people want
to go to and from the same places during a short period of time.
Rail systems are too expensive and are technologically unsuited
for low volumes of demand. Ordinary buses cannot maintain
sufficiently frequent service in outlying areas to attract any but
those who have no alternative. What is needed is a public
transit system which can respond dynamically to the needs of
these areas, that is, a system whose routes and schedules are
both flexible and ubiquitous.

The Dial-a-Bus, which is a hybrid between an ordinary
bus and a taxi, could be the basis for such flexibility. It
would pick up passengers at their doors or at a nearby bus stop
shortly after they have telephoned for service. The computer would
know the location of its vehicles, how many passengers were on
them, and where they were heading. It would select the right vehicle
and dispatch it to the caller according to some optimal routing program
which had been devised for the system. Thus, the system could
readily link many origins to many destinations.

A Dial-a-Bus, with it's position established by automatic vehicle
monitoring, can be routed by  computer and a communication link
to collect passengers who have called for service.

The diffused pattern of trip origins and destinations which
this system would serve is most dominant in low density suburbs.
But it also exists in a different form in the most thickly populated
urban areas.

The cost of taxi rides can be driven down by sharing
rides, and basically the Dial-a-Bus system is designed to
accomplish this. Data from the new systems study suggest that,
depending on demand, door-to-door transit can serve its
passengers almost as fast as a private taxi but at one-quarter
to one-half the price, indeed, at only slightly more than the fare
for a conventional bus.

With its operational flexibility, the Dial-a-Bus system could
be programmed to give different levels of service for different
fares. At one extreme it might offer unscheduled single pas-
senger door-to-door service, like a taxi, or multi-passenger serv-
ice, like a jitney. At the other extreme it might operate like a
bus service, picking up passengers along specified routes which
could include several home pick-ups. The system might also be
programmed to rendezvous with an express or line-haul carrier,
and in serving as either a collector or distributor, provide the
opportunity to improve the complete transportation service.

The major point is that the Dial-a-Bus might do what no
other transit system now does: Handle door-to-door travel
demand at the time of the demand. This means that the system
would attract more off-peak business than does conventional
transit. And if it does attract enough passengers, the off-peak
revenue would help Dial-a-Bus avoid the same financial prob-
lems of conventional transit, which is used heavily only 3 or 4
hours per day. It could also help reduce dependence upon
automobiles.

Technically, there is little question that the system will work.
Any number of existing vehicles can comfortably carry 12 to
24 passengers. Some of the best are now offering service to
airports. Present computers, radio communications, and telephone
links are fully adequate to the major needs of Dial-a-Bus.

Mathematical routing and the associated computer programming
present no real obstacles. What must be done is to put these
isolated elements together into a unified system. Dial-a-Bus
service could be made somewhat more efficient if the buses
were equipped with automatic monitors to report each vehicle's
location, to the dispatchers at frequent intervals. Although these
monitors do not now exist, there is no technological barrier to
developing them, as discussed above under the automatic ve-
hicle monitoring subsystem.

The cost for a given level of Dial-a-Bus service is a function
of many variables. These include the nature of the street system,
the cruising speed of the vehicle, the distribution of demand,
and the size of the area served. Perhaps the most uncertain of
these variables is demand density, the number of trips generated
per square mile per hour. Dial-a-Bus systems probably will be
most efficient at demand densities of 100 trips per square mile
per hour—a level that is barely practicable for conventional
bus service.

A limited demonstration of the Dial-a-Bus concept, using
existing equipment, could almost certainly be achieved within
3 years at a cost of less than $1 million. A definitive full-scale
demonstration of Dial-a-Bus service, using vehicles and control
equipment specifically designed for this purpose to test the full
range of possible benefits, probably could be completed within
7 years at a cost of less than $20 million.

http://faculty.washington.edu/jbs/itrans/tomtrans.htm