Can We Model Driver Perceptions? An In-Situ Experiment in Real-World Conditions

Abstract

It is clear that perceptions play a significant role in traveler decisions. Consequently, traveler perceptions are a corner stone in the feasibility of traveler information systems; for traveler information systems are only valuable if the drivers are incapable of accurately acquiring the provided information on their own, and if the provided information is relevant for the drivers' decision criteria. Accuracy of traveler perceptions has been repeatedly researched in public transportation, and has been found to vary according to different reasons. However, in spite of the clear significance of traveler perceptions, minimal effort has been put into modeling it. Almost all travel behavior models are based on traveler experiences, which are assumed to reflect traveler perceptions via the addition of some random error component. This works introduces an alternative approach: instead of adding an error component to represent driver perceptions, it proposes to model driver perceptions. This work is based on a real-world route choice experiment of a sample of 20 drivers who made more than 2,000 real-world route choices. Each of the drivers' experiences, perceptions, and choices were recorded, analyzed and cross examined. The paper demonstrates that: i) driver experiences are different from driver perceptions, ii) driver perceptions explain driver choices better than driver experiences, iii) it is possible to model and predict driver perceptions of travel distance, time and speed.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 2 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.