Human-Centered AI to Support an Adaptive Management of Human-Machine Transitions with Vehicle Automation

Bellet, Thierry; Banet, Aurélie; Petiot, Marie; Richard, Bertrand; Quick, Joshua · 2020 · Crossref

DOI: 10.3390/info12010013

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 challenge of managing Human-Machine Transitions (HMT) in vehicle automation by proposing a Human-Centered Artificial Intelligence (HCAI) algorithm. The research is motivated by the need to design Advanced Driving Aid Systems (ADAS) that adapt to individual driver capabilities and situational criticality, rather than relying solely on technological capabilities. The authors argue that effective automation requires monitoring driver behavior and real-time risks to ensure safe and accepted transitions between manual and automated control. The study employs a Human-Centered Design (HCD) methodology across five stages. First, the authors established theoretical foundations based on the HMT cycle, which defines interaction modalities (information, warning, takeover) and transition types. Second, they derived ergonomic requirements by analyzing a naturalistic dataset from 99 drivers, including 76 elderly drivers, 15 novice drivers, and 8 experienced drivers. This dataset, comprising 2800 km of driving, identified specific errors and difficulties, particularly during intersection crossings and highway lane changes. Third, an AI algorithm was developed to monitor these factors and manage HMI adaptively. Fourth, the algorithm was integrated into ADAS demonstrators on driving simulators for highway and urban scenarios. Finally, the system was evaluated through user tests with 30 participants divided into three groups of 10, representing positive, neutral, and reluctant attitudes toward vehicle automation. The results indicate that the HCAI algorithm effectively manages HMT by adapting to driver needs and situational risks. The evaluation confirmed the algorithm's efficiency and effectiveness in supporting transitions. Furthermore, the study found high levels of acceptance, perceived utility, usability, and attractiveness among all three participant groups. Drivers reported increased satisfaction with the adaptive nature of the system, which tailored interactions based on their specific profiles and the criticality of the driving situation. The significance of this work lies in its demonstration that Human-Centered AI can enhance the safety and acceptance of vehicle automation. By grounding the design in real-world driver errors and needs, the proposed approach offers a framework for developing adaptive systems that bridge the gap between manual and automated driving. The findings suggest that such systems can improve user experience and road safety by providing context-aware assistance, thereby addressing key barriers to the adoption of advanced driving aids.

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.

• automation
• acceptance adoption
• automation surprise

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: self report data
• Theoretical Contribution: computational model, conceptual framework