Comprehensive Modeling and Evaluation of Workload in Driving Simulation Using the VACP Paradigm

Abstract

Understanding and quantifying driver workload is essential for designing safe and effective human–vehicle interaction systems, especially in complex, multi-task driving contexts. Traditional workload measures often rely on aggregate or subjective indices, limiting insight into how distinct perceptual, auditory, cognitive, and motor demands evolve over time. To address this gap, this study adopts the Visual–Auditory–Cognitive–Psychomotor (VACP) paradigm as a structured framework for decomposing driver workload and evaluating its physiological validity using eye gaze measures. Using a pre-existing simulated driving dataset, workload was modelled during a primary driving task combined with three secondary tasks: braking, dialogue-based interactions, and a tactile Detection Response Task (DRT). The driving timeline was segmented into five conditions: baseline driving, braking only, dialogue only, DRT only, and simultaneous braking–dialogue events with onset asynchrony. For each condition, detailed VACP workload models quantified visual, auditory, cognitive, and psychomotor demands across task phases. Physiological relevance was assessed using pupillometry as an objective indicator of cognitive workload. Pupil diameter was analysed in relation to time-varying VACP workload estimates. Results showed a clear correspondence between increased VACP-defined workload and pupil dilation. Pronounced pupil responses occurred during high-demand braking and dialogue events involving concurrent workload components, while smaller but consistent responses were observed for discrete secondary tasks such as DRT and dialogue interactions. These findings demonstrated that pupil diameter is sensitive to both magnitude and composition of VACP workload, supporting the framework’s ability to capture meaningful variations in driver demand. Overall, the results validated the VACP paradigm as a systematic tool for modelling driver workload in complex, multi-task scenarios, with implications for driver monitoring, human–machine interface evaluation, and adaptive vehicle technologies.

Keywords: Visual Workload, Auditory Workload, Cognitive Workload, Psychomotor Workload, Human-computer Interaction, Workload Components.

DOI: 10.54941/ahfe1007524