Abstract

Flexible passenger transportation occurs in response to the rigidity of the routes and travel conditions offered by conventional public transportation systems. In this modality, both origin and destination are explicitly indicated by the passenger. Usually, the definition of travel routes is oriented towards operational efficiency, which is not necessarily related to the passenger’s travel experience. Therefore, the question arises regarding the extent of flexibility of passenger transportation solutions currently available in the industry. To the best of our knowledge, no proposals have incorporated artificial emotions in cognitive agents applied to the domain of flexible passenger transportation, allowing these agents to autonomously deliberate and decide after consideration of objective (travel time or cost) and subjective variables (emotions and satisfaction) within a unique integrated layer, with the help of passenger and travel independent profiles. This work considers a single hypothetical city map based on the use of a matrix of dimensions 500 × 500, which has congested, touristic, unsafe, and neutral streets. Three types of artificial agents were defined: Passenger Agent, Vehicle Agent, and Fleet Management Agent, with specific functions performed by each. To select a proposal, the Passenger Agent determines a selectivity index from the previous calculation of three metrics: travel time, travel cost, and utility. By considering thirty different experimental scenarios, each one executed 1000 times independently, it is possible to observe promising results that demonstrate high utility for the passenger in some cases, while the travel time and cost do not exceed the average of the system. © 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.