BROWSE
Detail
Conference Paper
Human Responses to Network Latency in Level 4 Autonomous Vehicle Teleoperation: From Perception–Control Breakdown to Trust Decay
- Authors
- Jiwoo Han, Woojin Kim, Eun Hye Jang, Mi Chang, Daesub Yoon, Yang Koo Lee, Kezia Amanda Kurniadi
- Issue Date
- 2026-04
- Citation
- Conference on Human Factors in Computing Systems (CHI) 2026, pp.1-5
- Publisher
- ACM
- Language
- English
- Type
- Conference Paper
- Abstract
- Network latency disrupts the perception–action loop in teleoperated Level 4 autonomous vehicles, compromising safety and eroding trust. QoS metrics can flag connectivity degradation, but they do not capture the operator’s cognitive difficulty or passengers’ anxiety. We propose a multimodal approach integrating vehicle telemetry with operator and passenger eye tracking and physiological signals to decode the human impact of network delay during cut-in scenarios. In a high-fidelity simulation study (N = 51), our model achieved robust delay classification (AUC = 0.8175) and uncovered behavioral markers across the loop. Under delay, operators exhibit a Perception–Control Breakdown marked by heightened perceptual uncertainty and over-correction, followed by Cognitive Reconfiguration in which scanning strategies simplify into rigid, lower-entropy patterns. Passengers show evidence of Trust Decay and an Attentional Lock-in effect during recovery. We conclude with latency-aware interface guidelines including predictive displays, intent-level interaction, event boundary visualization, and explainable feedback to restore the human connection when the network falters.
- Keyword
- Teleoperation, Autonomous Vehicles, Network Latency, Eye-tracking, Physiological Computing, Situation Awareness, Passenger Trust
- KSP Keywords
- Autonomous vehicle, Boundary visualization, Cognitive difficulty, High-fidelity simulation, Interface guidelines, Lock-in effect, Multimodal approach, Network Delay, Network latency, Physiological signals, QOS Metrics
This work is distributed under the term of Creative Commons License (CCL)
(CC BY)
(CC BY)
