Evaluation of a Natural Speech Based Informational Inquiry System as a Potential Means to Increase Transit Utilization
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 evaluates the potential of natural speech-based inquiry systems to increase public transit utilization and assesses the cognitive workload of voice interfaces in driving scenarios. The research was motivated by the challenge of encouraging non-users to develop familiarity with transit systems. However, rapid advancements in third-party transit applications made long-term evaluation of a specific speech interface difficult, leading the project to broaden its scope to examine speech-based interfaces across various transportation contexts. The study comprised two primary components. First, researchers expanded the CityBrowser spoken dialogue prototype to include transit information, utilizing crowdsourcing via Amazon Mechanical Turk to collect queries for training the system’s semantic understanding. Although initial attempts to incorporate real-time Boston transit data proved unreliable, subsequent versions integrated directions via a Google API. Second, the project conducted an in-vehicle assessment using a 2010 Lincoln MKS equipped with a SYNC™ voice interface. Sixty subjects, evenly distributed by gender and across two age groups (20–29 and 60–69), performed six tasks: manual and voice-controlled radio tuning, navigation destination entry, MP3 song selection, phone dialing, and an auditory n-back calibration task. The vehicle was instrumented to record synchronized data from the controller area network, physiology monitors (heart rate and skin conductance), eye-tracking, and cameras. The findings indicate that voice interfaces offer both benefits and cautions regarding driver workload. Compared to manual radio tuning, which served as a high-demand baseline, voice-controlled radio tuning resulted in lower subjective workload, less compensatory speed reduction, and lower physiological arousal. Similarly, voice-based destination entry produced lower subjective workload and arousal than manual entry, supporting its use as an acceptable method for navigation input while driving. Conversely, selecting a pre-set radio station via voice command yielded higher subjective workload and physiological arousal than manual operation. The study also evaluated the n-back task as a cognitive workload metric, suggesting that the 1-back level, rather than the 2-back level, may serve as a more suitable threshold for acceptable cognitive demand in in-vehicle systems. The significance of this work lies in its empirical assessment of human-machine interaction in transportation. The results suggest that while voice interfaces can reduce cognitive load for complex tasks like destination entry, they may increase workload for simpler tasks like station selection. These findings provide critical insights for designing safer and more effective driver-vehicle interfaces, balancing the potential benefits of hands-free operation against the risks of increased cognitive distraction.
Key finding
Voice-based destination entry produced lower subjective workload, less compensatory speed reduction, and lower physiological arousal than manual hard radio tuning, though selecting a preset radio station by voice was rated higher in workload than doing so manually.
Methodology
on_road
Sample size: 60
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 | success | — | — | — | 3 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
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
- Empirical Findings: self report data, observational prevalence