Quantifying the Influence of Image Quality on Operator Reaction Times for Teleoperated Road Vehicles

Hoffmann, Simon; Willert, Felix; Hofbauer, Markus; Schimpe, Andreas; Diermeyer, Frank · 2022 · Crossref

DOI: 10.54941/ahfe1002322

archive: archived pipeline: cataloged verified

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

Summary

This study investigates the impact of image quality (IQ) on operator reaction times in teleoperated driving (ToD), a concept used to resolve edge-case situations for automated vehicles when they exit their operational design domain. Because ToD relies on video streams transmitted over cellular networks, dynamic compression adjustments often degrade image quality, potentially hindering an operator’s situational awareness and safety. The authors hypothesized that lower IQ significantly increases operator reaction times to dynamic obstacles, thereby compromising safety. To test this, the researchers conducted a within-subject user study with 34 participants. The experimental scenario involved a ball rolling onto the road from an occlusion, chosen for its small size and reproducibility. Participants viewed 20 recorded videos, each encoded at five different quality levels (Q20–Q50) using the x264 encoder with a "constant quality" setting to ensure defined quality levels independent of bitrate fluctuations. The study utilized a randomized order to mitigate learning and sequence effects. Objective data included reaction times measured from the appearance of the obstacle to brake input, while subjective data comprised numerical ratings of image quality, perceived influence on task performance, and sufficiency for ToD. The results demonstrated a significant influence of IQ on both subjective and objective task performance. Subjectively, participants rated IQ and its impact on performance significantly differently across quality levels, with medium to large effect sizes. Objectively, reaction times increased significantly as image quality decreased. Specifically, reaction times at the lowest quality level (Q50, mean 939 ms) were significantly higher than at higher quality levels (e.g., Q44 at 729 ms, Q36 at 607 ms). Notably, while subjective ratings detected significant differences between Q36 and higher quality levels, objective reaction times did not show a significant difference between Q36 and the highest quality levels (Q20/Q28). This suggests that operators’ subjective judgments of quality degradation are more conservative and sensitive than their actual motor reaction times in this specific scenario. The study concludes that image quality critically affects teleoperation safety by increasing reaction times, particularly at lower quality levels. The findings indicate that subjective assessments may serve as a more sensitive indicator of quality degradation than objective reaction times for this task. However, the authors note limitations, including the use of a single dynamic obstacle scenario and the fact that participants were not actively driving, which may affect the generalizability of absolute reaction times. Future work should investigate static obstacles and integrate IQ metrics that correlate with task performance to establish standardized quality thresholds for safe teleoperation.

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.

• teleoperation remote driving

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: behavioral performance data