Adaptive driver modelling in ADAS to improve user acceptance: A study using naturalistic data

Fleming, James M.; Allison, Craig K.; Yan, Xingda; Lot, Roberto; Stanton, Neville A. · 2018 · OpenAlex-citations

DOI: 10.1016/j.ssci.2018.08.023

archive: archived pipeline: cataloged verified

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

Summary

This study addresses the challenge of improving user acceptance for Advanced Driver Assistance Systems (ADAS) by developing adaptive models that respect individual driving styles. While ADAS technologies like collision and curve warnings significantly reduce accident rates, their efficacy is often limited by driver disablement due to high false alarm rates. The authors argue that current driver behavior models are too prescriptive and fail to account for the wide variability in individual driving habits. To bridge this gap, the paper evaluates existing car-following and cornering models using naturalistic driving data, aiming to identify a small set of measurable parameters that can characterize driver behavior in real-time. The research utilized data from a naturalistic driving study conducted at the University of Southampton. Data was collected using a portable, non-intrusive device called ADAM, which recorded GPS position, velocity, acceleration, and inter-vehicle spacing via stereo cameras. The study involved nine participants: six for car-following analysis in urban traffic and three for cornering analysis in rural conditions. The authors filtered the time-series data to isolate steady-state driving conditions, such as constant speed following and curve negotiation, to validate theoretical models against real-world observations. The findings reveal significant limitations in existing driver models. In car-following scenarios, the data showed a wide range of acceptable inter-vehicle spacings, contradicting models that assume a single preferred spacing. Instead, the authors identified a lower bound on spacing that increased linearly with speed, characterized by minimum distance ($s_{min}$) and minimum time headway ($T_{min}$) parameters. For cornering, the study validated the lateral acceleration bounds proposed by Reymond et al., finding that drivers maintain an upper limit on lateral acceleration that decreases with speed. However, the authors noted that these bounds were influenced by UK roadway design codes. Alternative models, such as the "two-thirds law" and power-law relationships, were found to overestimate allowable lateral acceleration at specific speeds. The significance of this work lies in its proposal for adaptive ADAS design. By identifying specific parameters—such as minimum time headway and lateral acceleration limits—that can be estimated in real-time, the study suggests a pathway for systems to adapt to individual drivers. This adaptation aims to reduce false alarms, thereby improving user acceptance and ensuring the safety benefits of ADAS are realized. Furthermore, the framework supports the development of eco-driving assistance systems by providing accurate baselines for normal driving behavior, allowing for more effective feedback on fuel-efficient practices.

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.

• speed choice
• following distance
• behavioral adaptation risk compensation
• steering pattern
• eco driving
• situational awareness

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: behavioral performance data
• Methodological Resource: dataset resource
• Theoretical Contribution: computational model