Driver Engagement

Deiana, Francesco; Roig, Adria; Jackson, James; Periago, Cristina; Cabuti Ferrer, Clara · 2023 · Crossref

DOI: 10.54941/ahfe1003821

archive: archived pipeline: cataloged

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

Summary

**Driver Engagement** This paper addresses the critical challenge of ensuring driver engagement in Level 2 (L2) advanced driver assistance systems (ADAS), where drivers remain responsible for vehicle control. As L2 systems become prevalent, concerns have arisen regarding driver misuse and lowered attention levels, necessitating validation methods that assess how system design influences driver vigilance and reaction capabilities. The study aims to develop and pilot a test methodology for an automotive proving ground that quantifies driver engagement through both subjective and objective measures. The methodology involved a between-subjects pilot test with 39 naïve drivers (20 male, 20 female, with one participant unaccounted for in the gender split or a typo in the source, though the text states 20/20 which sums to 40, likely indicating a minor inconsistency in the text, but the total n=39 is stated). Participants drove two distinct vehicles: a Volkswagen Golf 8 (medium L2) and a Tesla Model 3 (advanced L2), both instrumented with CAN bus data, GPS, and video cameras. The test consisted of a 40-minute drive on a 2.7 km highway loop, including a 10-minute familiarization phase without assistance and a 30-minute phase with adaptive cruise control and lateral assistance active at 60 km/h. Subjective data were collected every five minutes using the Integrated Workload Scale (IWS) and the Trust in Automated System Survey (TASS). Objective data focused on Time To Collision (TTC) during a critical "cut-out" event where a lead vehicle abruptly moved to reveal a stationary obstacle, requiring the driver to intervene. Results indicated significant differences in driver engagement between the two systems. The medium L2 system (Golf 8) was associated with higher perceived mental workload (mean values ranging from 1.75 to 2.10) compared to the advanced L2 system (Tesla Model 3, mean values ranging from 1.20 to 1.75). Conversely, trust levels were higher for the advanced system, with a maximum mean trust of 96.1% versus 93% for the medium system. Objectively, driver vigilance was lower in the advanced system; 60% of participants in the Tesla Model 3 had a TTC below the 1.5-second safety threshold, compared to only 45% in the Golf 8. This suggests an inverse relationship between perceived trust and mental workload, where higher trust in the advanced system led to reduced vigilance and slower reaction times to the emergency scenario. The study concludes that the type of L2 system significantly influences driver behavior, with advanced systems potentially fostering over-trust that compromises safety. The proposed methodology provides a replicable platform for assessing driver engagement in controlled environments, with implications for system design verification, consumer testing, and regulatory standards. Future work will refine the protocol, include braking-time analysis, and expand to Level 3 systems and demographic clustering.

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 qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

Topics

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

• hands on hands off engagement
• automation
• situational awareness
• automation surprise
• temporal
• trust calibration

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
• Methodological Resource: validation psychometrics
• Theoretical Contribution: conceptual framework