Empirical analysis of an in-car speed, headway and lane use Advisory system

Empirical analysis of an in-car speed, headway and lane use Advisory system

Schakel, W.J.
Van Arem, B.
Van Lint, J.W.C.

Civil Engineering and Geosciences

Transport & Planning

2014-11-13

For a recently developed in-car speed, headway and lane use advisory system, this paper investigates empirically advice validity (advice given in correct traffic circumstances), credibility (advice logical to drivers) and frequency. The system has been developed to optimize traffic flow by giving advice on a tactical scale. This scale allows traffic flow improvement and fills a gap as most ITS which aim to optimize traffic flow operate on the operational or strategic scale. Using log files of the actual system for a period of two weeks, the validity, credibility and frequency of advices is determined. Validity is not guaranteed as the advices are determined based on a predicted traffic state due to data delay and as the advices are based on the predicted traffic state 1 minute in the future. Given that the advisory system is a first in its kind, a new methodology was developed to assess the system, based on the use of virtual trajectories and defining indicators to assess, validity, credibility and frequency. The analysis shows that many advices are indeed valid and credible, but some are not, allowing room for improvement. Advice frequency is found to be reasonable. The analysis also shows that in-car filtering of advices, i.e. merging equal advices, is important to lower the frequency.

http://resolver.tudelft.nl/uuid:8b79d7c7-9c69-462e-bcb4-2c8ca50f563a

TRAIL

2nd TRAIL Internal PhD Conference, Delft, The Netherlands, 13 November 2014; Authors version

Institutional Repository

conference paper

(c) 2014 The Author(s)