A New Measure of Travel Time Reliability for In-Vehicle Navigation Systems

Kaparias, Ioannis; Bell, Michael G.H.; Belzner, Heidrun · 2008 · Crossref

DOI: 10.1080/15472450802448237

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the need for a practical travel time reliability measure suitable for implementation in the dynamic routing algorithms of in-vehicle navigation systems. Existing reliability metrics, such as the probability of uncongested travel, are often unintelligible to drivers or fail to translate into expected delay values. Furthermore, probability-based measures are biased by route length, penalizing longer routes regardless of actual congestion risk. The authors aim to define a new measure that reflects the maximum possible delay, is dimensionless, independent of link length, and expressed on a [0,1] continuum to integrate seamlessly with existing routing algorithms. The methodology assumes that travel times follow a log-normal distribution, consistent with empirical evidence of right-skewed travel time data. The authors derive two reliability indices: a lateness index ($r_L$) and an earliness index ($r_E$). These indices represent the ratio of the mean travel time to the upper and lower bounds of the travel time confidence interval, respectively. Because travel time data is often unavailable in practice, the authors perform mathematical transformations to express these indices in terms of speed distributions. Using Taylor series expansions and relationships between space-mean and time-mean speeds, they derive expressions for the travel time variation logarithm based on speed variance and mean speed. Additionally, the paper derives a formula for calculating the reliability of an entire route based on the statistical properties of its constituent links, accounting for the aggregation of mean travel times and variances. The study finds that the proposed indices effectively quantify reliability on a standardized scale. The lateness index is identified as the critical metric due to the right-skewed nature of the log-normal distribution, where the upper bound deviation from the mean is larger than the lower bound. The derived mathematical relationships allow navigation systems to compute reliability using readily available speed statistics rather than requiring complex travel time distributions. The accuracy of the measure is assessed through a field experiment, where estimated travel times derived from the new reliability indices are compared against actual observed travel times. The results demonstrate that the measure accurately reflects travel time variability and provides a robust basis for dynamic routing. The significance of this work lies in its practical application to intelligent transportation systems. By providing a driver-comprehensible, dimensionless, and length-neutral reliability metric, the paper enables navigation systems to offer more accurate route guidance that accounts for travel time uncertainty. This advancement allows drivers to make informed decisions based on the likelihood of delays, improving the utility of dynamic routing functions. The ability to calculate route reliability from speed data ensures the measure is feasible for real-world implementation in both autonomous and supported navigation architectures.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• exposure measurement