Understanding Automation Surprise in Non-Critical Highly Automated Driving: An Initial On-Road Probing Study
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 phenomenon of automation surprise (AS) in non-critical, highly automated driving scenarios, shifting the focus from safety-critical failures to user experience and comfort. As automated vehicles (AVs) advance to SAE levels 3–5, users increasingly engage in non-driving related activities (NDRAs), diverting attention from the road. The authors argue that AS—defined as a cognitive-emotional response to unexpected system behavior—can occur even in routine, non-emergent conditions, potentially causing discomfort and reducing user acceptance. Motivated by a gap in literature that primarily treats AS as a safety or operational error rather than a design challenge for user experience, the study aims to expand the understanding of AS to include these daily, non-critical interactions. To investigate this, the researchers conducted an initial on-road probing study using a qualitative approach. They employed a Wizard of Oz (WoZ) setup in a modified BMW i3, where a human driver concealed behind a partition simulated automated driving behaviors. Three participants, all students with Dutch driver’s licenses, drove for approximately 40 minutes on a predefined route in Eindhoven, Netherlands, encompassing various conditions such as city traffic, highways, and roundabouts. Participants were instructed to engage in NDRAs using their phones while employing the think-aloud method to provide continuous feedback on their cognitive and emotional states. Data were analyzed through transcription and annotation of audio recordings, mapping instances of surprise and related emotions onto the driving route to identify contextual triggers. The results indicate that automation surprise is not limited to high-speed or critical scenarios but frequently occurs during slow-moving city traffic, smooth deceleration, or prolonged steady driving. Participants experienced surprise triggered by unexpected vehicle movements, such as abrupt lane changes, sudden braking, or unanticipated acceleration, as well as by confusing interface cues like turn signals activated without immediate action. These events elicited immediate emotional responses including fear, anxiety, stress, and confusion, alongside cognitive efforts to understand the vehicle’s reasoning. However, the study found that surprise could be mitigated through familiarity; repeated exposure to specific scenarios reduced the surprise effect, and participants often calmed down once they visually confirmed the reason for the vehicle’s action, such as an upcoming traffic light. The significance of these findings lies in the proposal of an expanded definition of automation surprise as a complex phenomenon arising when users fail to grasp the AV’s actions, predict responses to road conditions, or encounter abrupt events. The authors conclude that AS in non-critical conditions significantly impacts user experience and trust, highlighting the need for human-centered design strategies that improve communication between the system and the user. By understanding the cognitive and emotional dimensions of AS, future research can develop mitigation strategies, such as better interface explanations and adaptive behaviors, to enhance user comfort and acceptance of automated driving technologies.
Key finding
Automation surprise occurs in non-critical driving conditions when unexpected vehicle behaviors or signals trigger cognitive confusion and emotional stress, which can be mitigated by user understanding of the vehicle's actions.
Methodology
on_road
Sample size: 3
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 scout_discovery on 2026-05-08.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | partial | scout | — | — | 2 | 2026-05-08 |
| archive | success | canonical_url | — | — | 4 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-08 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | partial | — | — | — | 2 | 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.
- automation surprise
- automation
- acceptance adoption
- situational awareness
- automation complacency bias
- 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
- Theoretical Contribution: conceptual framework