Abstract
Using a private car is a transportation system very common in industrialized countries. However, it causes different problems such as overuse of oil, traffic jams causing earth pollution, health problems and an inefficient use of personal time. One possible solution to these problems is carpooling, i.e. sharing a trip on a private car of a driver with one or more passengers. Carpooling would reduce the number of cars on streets hence providing worldwide environmental, economical and social benefits. The matching of drivers and passengers can be facilitated by information and communication technologies. Typically, a driver inserts on a web-site the availability of empty seats on his/her car for a planned trip and potential passengers can search for trips and contact the drivers. This process is slow and can be appropriate for long trips planned days in advance. We call this static carpooling and we note it is not used frequently by people even if there are already many web-sites offering this service and in fact the only real open challenge is widespread adoption. Dynamic carpooling, on the other hand, takes advantage of the recent and increasing adoption of Internet-connected geo-aware mobile devices for enabling impromptu trip opportunities. Passengers request trips directly on the street and can find a suitable ride in just few minutes. Currently there are no dynamic carpooling systems widely used. Every attempt to create and organize such systems failed. This paper reviews the state of the art of dynamic carpooling. It identifies the most important issues against the adoption of dynamic carpooling systems and the proposed solutions for such issues. It proposes a first input on solving the problem of mass-adopting dynamic carpooling systems.