Improving carpool flexibility without compromising trust or guaranteed rides

From IFORS Developing Countries OR Resources

Jump to: navigation, search

by: Fernando Lobo Pimentel


Aiming at the same transport with less car usage, carpool systems are a more economical and ecological way to travel when compared to drive-alone behaviour. However, carpooling is difficult to promote. People do not carpool for schedule flexibility and trust concerns, and when addressing the first issue by bringing more people to the carpools to increase schedule options, one loses on the trust side because prior acquaintance is no longer guaranteed, or one looses a guaranteed ride. We tried to address this problem.

We used operations research (OR) methodology to formulate the problem, then we relaxed a carpool system's restriction: the schedule coincidence requirement. This allowed the design of a several departure time carpool. Next we optimized the system using LP (linear programming) for a sub problem. Finally two preliminary surveys were conducted. The first in Oeiras municipality (Portugal), to test potential carpoolers adherence to the model, and the second through an email chain, to anticipate effective enrolment in a carpool with a specific several departure time schedule.

We found that a small group of people with different but compatible schedules, and who meet each other previously, can join the same carpool and benefit from it, provided that the system operates under the optimised configurations presented, so that the increase in the number of departure times available does not decrease vehicle occupancy rates more than necessary. Surveys revealed the likely readiness and schedule compatibility of about 10% of car commuters to enrol in such system.

By designing the carpool system around the idea of several guaranteed departures available, groups can be smaller and steadier because they self-contain wider schedule options. Scale increases are no longer mandatory and riding with strangers or unpredictable ride availability can be prevented. Therefore our model is a contribution to improve carpool flexibility without compromising trust concerns or guaranteed rides.

Keywords: Carpooling, Carpool flexibility, Operations Research, Optimization

Link to material:

Personal tools