Session: 07-01-02: Injury Sensing and Mitigation

Paper Number: 172804

A Predictive Vibration-Based Driver Fatigue Model 

INTRODUCTION

Exposure to whole-body vibrations significantly impacts driver performance by increasing mental and physical fatigue. Physical fatigue results from excessive or extended physical exertion and is related to reduced contractile performance and inadequate activation of muscles, resulting in reduced performance. In comparison, mental fatigue is more abstract but can be observed through changes in mental state (drowsiness) and reaction time as well as through physiological measures, such as electroencephalography (EEG). Fatigue can be influenced by a variety of factors, such as one’s physical health, nutrition, the environment, and the length of exertion. Being fatigued can greatly impact a driver’s ability to respond to critical driving events in time.

 

There are multiple methods to detect fatigue while driving, such as monitoring physiological activity using EEG or ECG and observing physical characteristics (facial features and eye movement) through visual tracking, but they cannot be used in a predictive capacity. Therefore, we have developed a reduced-order model that captures human fatigue development due to both cumulative effects of instantaneous frequency-weighted vibrations, and then predicts potential development of vehicle occupant fatigue. Existing theories and models were integrated and adapted to achieve this combined fatigue prediction capability.

 

METHODS

The model described here predicts fatigue based on single or multiple vibration profiles., These can be recorded from physical experiment, obtained from vehicle dynamics simulation, or generated randomized vibration profiles based on terrain roughness and vehicle parameters. The vibration profiles are used to calculate vibration dose value (VDV) and absorbed power as metrics of fatigue development. Further, the fatigue model incorporates components for rest parameters, driving duration, and vibration intensity to calculate driving fatigue, as well as to detect instantaneous fatigue resulting from vibrations at frequencies known to result in premature fatigue or drowsiness. To accomplish these tasks, the terrain and acceleration, general driver fatigue, and vibration-based driver fatigue are considered.

 

The acceleration response can be used directly, as calculated from a terrain profile and vehicle dynamics simulations or it can be generated based on terrain roughness obtained from the International Roughness Index terrain description via a non-stationary Laplace vibration terrain generator. After terrain profiles are read or generated, they are converted to acceleration response via an eight degree-of-freedom full car model which includes a driver seat characterization.

 

The acceleration response is then used to calculate the vibration exposure as determined by calculating established vibration metrics, including the vibration dose value (VDV) and absorbed power (AP). VDV gives a total vibration dose based on experienced accelerations, and full details for calculating it are included in the ISO 2631 standard. Comparatively, AP measures the energy absorption rate at specific frequencies, with a focus on the resonant frequencies of internal organs.

 

The present model implements two methods of fatigue calculation. The first method calculates frequency-based fatigue based on identifying dominant frequencies and comparing them to specific frequencies that have been found to induce acute fatigue, specifically 1.8, 3, and 6 Hz. To start, the vibration acceleration profile must be converted from the time domain to the frequency domain, which is accomplished using Fourier transforms. This transformation is performed and examined on 20-minute intervals, based on the time that acute fatigue has been shown to occur, and is only performed if a minimum defined threshold is exceeded. The second method is based on a fatigue model where cumulative driver fatigue is based on drive duration and a person’s amount of sleep, rest, and circadian rhythm.

 

RESULTS AND CONCLUSIONS

The model described here calculates a driver’s vibration exposure and predicts the potential to experience driving or acute frequency-based fatigue from vibrations produced by acceleration over time. These predictions are based on a combination of vibration exposure metrics, including duration, intensity, and experienced vibration frequencies, as produced through a simulated vehicle and driver condition. This model will help identify potential driver fatigue vehicle-scenario combinations that increase the risk of costly and dangerous accidents.

Presenting Author: Raheleh Miralami Mississippi State University

Presenting Author Biography: Raheleh (Rahel) Miralami is an Assistant Research Professor at the Center for Advanced Vehicular Systems at Mississippi State University. With a B.S. and M.S. in Mechanical Engineering and a Ph.D. in Biomedical Science, she brings a strong interdisciplinary foundation to her research and teaching. Her work spans biomechanics, mechanobiology, biomaterials, and bio-inspired design, integrating computational modeling and experimental biomechanics across multiple scales. Driven by a passion for innovation, Rahel aims to bridge the gap between theory and practice in regenerative medicine, ultimately translating scientific insights into improved clinical outcomes for patients with musculoskeletal conditions.

Authors:

Michael Murphy Mississippi State University
Raheleh Miralami Mississippi State University
Bohumir Jelinek Mississippi State University