Introduction
As autonomous vehicle (AV) technology advances, the transition of control from the vehicle to the human driver—often termed the "hand-off" problem—remains a critical safety challenge. In Level 3 autonomy, drivers frequently disengage from the primary driving task, leading to reduced situational awareness. When the system encounters a boundary condition, such as a sharp curve, traditional auditory or visual alerts can startle the driver, potentially resulting in erratic maneuvers. This study explores whole-body haptic feedback as a more intuitive method for maintaining driver engagement and facilitating safer transitions.
Methodology
The research utilized a high-fidelity Force Dynamics 401cr motion simulator, a 6-degree-of-freedom system capable of simulating realistic centrifugal forces and road vibrations. The experimental design compared two primary conditions: a baseline (no motion feedback) and a haptic feedback condition (whole-body motion cues).
To capture the driver's physiological and cognitive state, a multimodal sensor suite was employed:
- Eye-Tracking: Tobii Pro-Glasses 2 (60 Hz) measured gaze behavior and pupil diameter.
- EEG: A B-Alert X24 system (24 channels, 256 Hz) monitored cognitive load and cortical activation.
- AOI Analysis: Areas of Interest (AOI) were defined around upcoming road curves to quantify visual attention.
Results
Preliminary data from the initial participants indicate that whole-body haptic feedback significantly enhances driver engagement. Quantitative analysis of eye-tracking data revealed a marked increase in visual attention toward upcoming hazards when haptic cues were present.
As shown in the table above, the time spent fixating on curves increased from 4.49 seconds to 6.83 seconds ($p=0.002$). Furthermore, the duration of individual fixations increased ($3.45s$ vs $3.04s$, $p<0.001$), suggesting more thorough processing of road geometry.
Discussion: Cognitive and Physiological Impact
The physiological data supports the hypothesis that haptic feedback acts as a "priming" mechanism. Mean pupil diameter—a known proxy for cognitive effort and arousal—increased from 27 to 35 pixels ($p=0.008$) under the haptic condition. This suggests that the physical sensation of the car's movement effectively "wakes up" the driver's brain without the jarring effect of a traditional alarm.
EEG analysis further corroborated these findings. Increased Theta power in the frontal lobe indicates heightened attention and decision-making readiness. Additionally, activity in the C3 and C4 channels suggests motor cortex priming, preparing the driver for the physical requirements of steering and braking before the manual takeover occurs.
Conclusion and Future Work
This preliminary study demonstrates that whole-body haptic feedback significantly improves situational awareness and cognitive engagement during curve negotiation. By providing a continuous stream of physical information that mimics real-world physics, the system helps drivers build a mental model of the road ahead.
Future research will expand the sample size to ensure demographic diversity and explore more complex driving scenarios, such as compound curves and environmental hazards. The ultimate goal is the development of an adaptive haptic system that modulates feedback intensity based on real-time cognitive load.
Summary updated