A User-Centered Approach to Adapt the Human-Machine Cooperation Strategy in Autonomous Driving

Coeugnet, Stéphanie; Mars, Franck; Bueno, Mercedes; Sentouh, Chouki; Popieul, Jean-Christophe; Koustanaï, Arnaud; Pauzié, Annie; Tattegrain, Hélène · 2021 · Crossref

DOI: 10.1007/978-3-030-74608-7_73

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 designing safe and effective human-machine interfaces (HMI) for Level 3 autonomous vehicles, where drivers must act as fallbacks during critical situations. The research aims to adapt the cooperation strategy between the driver and the vehicle based on the driver’s real-time state, ensuring that monitoring systems are simple, non-intrusive, and provide necessary information without causing cognitive overload. The study employed a multi-phase, user-centered approach. Initial steps involved six focus groups and large-scale questionnaires with 2,619 participants to assess acceptance and information needs. Subsequent driving simulator studies analyzed visual, postural, and physiological indicators to model driver states, such as "out-of-the-loop" phenomena and emotional distractions. These findings informed the design of an adaptive HMI and a driver monitoring system, which were implemented in a Wizard of Oz (WOz) vehicle. The final assessment involved 52 drivers (mean age 38.5) on public roads. Participants engaged in non-driving tasks while the system simulated automated driving and issued planned (45-second budget) or unplanned (8-second budget) take-over requests. Data on visual behavior, posture, and physiology were recorded, followed by semi-directive interviews to evaluate system acceptance and feedback preferences. Results indicated strong support for adaptive monitoring: 97.7% of participants believed the system should assist drivers based on their state, and 86.1% were willing to be monitored, provided the technology was unobtrusive (e.g., integrated into the steering wheel). Regarding feedback modalities, 44% preferred a combination of visual and auditory alerts, noting that visual-only alerts were insufficient for drowsiness. However, 27.9% favored visual-only modes to avoid disturbing passengers. When presented with alert types, 61.9% preferred specific pictograms identifying the exact issue (e.g., hands-off, eyes-off) to facilitate correction, although some users expressed concern about information overload or irritation from frequent warnings. The study concludes that driver monitoring systems are essential for Level 3 automation but must balance safety gains with user comfort to ensure acceptance. The authors emphasize that systems should provide specific, actionable information during critical events while limiting the frequency and volume of data to prevent counterproductive effects. This user-centered design approach offers a framework for developing safer, more efficient cooperative driving systems that align with future users' expectations and needs.

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
• mode awareness
• automation surprise
• ehmi external hmi
• situational awareness
• odd communication

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).

• Applied Guidance: design guidelines
• Methodological Resource: tool software
• Theoretical Contribution: conceptual framework