Spoken Dialog Planning to Reduce User Distraction in Mobile Environments
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This report summarizes research conducted by the New England University Transportation Center, led by James Glass and Bruce Mehler at MIT, aimed at reducing user distraction in mobile environments, specifically focusing on drivers and pedestrians. The study addresses the safety hazards associated with speech-based interfaces in vehicles. While voice interfaces are intended to minimize visual and manual distractions, they impose cognitive loads that can detract from driving performance. The research was motivated by the need to design human-computer communication flows that minimize this cognitive burden, particularly as the National Highway Traffic Safety Administration considered new distraction guidelines for voice-activated controls. The project employed a multi-modal assessment approach, evaluating driving performance, visual engagement, and physiological indices of workload such as heart rate and skin conductance. The experimental design involved 80 drivers randomly assigned to operate either a 2013 Chevrolet Equinox or a Volvo XC60. The study compared two embedded vehicle systems: the Chevrolet MyLink, which allowed task completion via single voice commands, and the Volvo Sensus, which required multiple commands to navigate menu structures. Additionally, the research compared these embedded systems against portable technologies, including the Samsung Galaxy S4 smartphone and Google Glass, to assess trade-offs in demand between manual and voice interfaces. A separate study examined the effects of an “Expert Mode” in voice command systems, which streamlined tasks by removing confirmatory dialogue steps. The findings revealed that while voice interfaces reduced visual demand relative to visual-manual interfaces, neither embedded voice system entirely eliminated visual engagement. Drivers using the Chevrolet MyLink experienced greater reductions in visual demand during contact calling compared to Volvo Sensus users, but MyLink exhibited significantly higher error rates during destination address entry. When comparing embedded systems to smartphones, both embedded voice interfaces resulted in less eyes-off-road time than manual smartphone use, though the smartphone and MyLink allowed for shorter eyes-off-road times during address entry via compound voice commands, albeit with increased errors. Physiological measures indicated increased demand during secondary tasks compared to baseline driving, but no significant differences were found between smartphone and embedded systems. Furthermore, the “Expert Mode” significantly reduced task completion time but did not appreciably reduce visual engagement, as drivers continued to glance off the road for durations consistent with default modes. The significance of this work lies in its demonstration that modern voice-initiated systems are best understood as auditory-vocal-visual-manual-cognitive interactions rather than purely auditory tasks. The results imply that comprehensive demand assessment mechanisms must account for the time course of these mixed-mode tasks. The findings provide critical insights for interface design, suggesting that while voice commands can reduce certain types of distraction, they do not eliminate visual or cognitive demands, and design choices involving menu complexity versus command simplicity involve distinct trade-offs between error rates and visual engagement.
Key finding
Across 80 drivers, embedded voice interfaces reduced visual demand relative to visual-manual controls when calling a contact but never fully eliminated eyes-off-road time, and a streamlined Expert mode cut task time without reducing visual engagement.
Methodology
on_road
Sample size: 80
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. Discovered via bulk_ingest_rosap on 2026-05-23 (7 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | rosap | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 19 | 2026-06-11 |
| verify | partial | — | — | — | 3 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified_with_issues.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
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
- Theoretical Contribution: conceptual framework, theory or model