Vehicle trajectory prediction plays a pivotal role in enhancing advanced driver assistance systems (ADAS) and autonomous vehicles (AVs), crucial for collision avoidance, path planning, and traffic management. Traditional models often fail to account for variations in driver behaviour, such as eye movement patterns, which can substantially influence trajectory predictions. Our research presents an advanced trajectory prediction model that integrates Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks with both eye movement data and traditional vehicle dynamic information. The accuracy of these models was evaluated in a simulated environment designed to mimic real-world driving conditions, capturing extensive data on vehicle dynamics, including position, rotation, acceleration, speed, and eye movement patterns. Data collection was rigorously conducted with 17 drivers, each using a driving simulator that ran the Euro Truck Simulator software. The models were implemented and validated using Python 3.9 and Google Colab, chosen for their effectiveness in handling deep learning tasks. Our findings demonstrate that the inclusion of eye movement data alongside vehicle dynamics enhances the accuracy of trajectory predictions, significantly reducing both Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) as well as Mean Absolute Error Percentage (MAPE), compared to models based solely on vehicle dynamics. This improvement not only bolsters the precision of trajectory predictions for ADAS and AV systems but also significantly elevates their safety and operational efficiency.

Abstract Traffic accidents are a major global health and economic concern. As such, research into understanding driving behaviors becomes essential to minimize the associated risks. Among various factors that can influence driving behaviors, listening to music while driving is a complex task that needs investigation. Music can enhance arousal and manage stress, which could potentially improve one’s driving performance. Listening to music, however, also competes for cognitive resources, increasing one’s mental workload and potentially degrading the driving ability. This study investigates the impact of listening to music with different tempos on drivers and their performance using an Artificial Intelligence (AI) approach. A total of 26 participants are subjected to three driving scenarios in a unique simulated experiment, while utilizing a motion platform. The conditions are driving while listening to slow-tempo music, and fast-tempo music, and with no music. A dataset is created by collecting data through Tobii eye-tracking glasses, Equivital sensor belts, and a software tool. This dataset is preprocessed and used to train a convolutional neural network-long short-term memory (CNN-LSTM) model. This model’s performance is optimized through hyperparameter tuning and Chi-squared feature selection, in order to maximize accuracy and minimize computation time. The model performance is compared with those from a densely layered deep learning model and several classical machine learning models. The devised CNN-LSTM model outperforms other machine learning models, achieving an average accuracy rate of 99.29% with minimal variance across multiple evaluations, demonstrating its effectiveness and consistency in classifying drivers’ behaviors under varying auditory conditions.

Vehicle trajectory prediction is essential in ensuring the safe and efficient operation of advanced driver assistance systems (ADAS) and autonomous vehicles (AVs), as it enables highly efficient collision avoidance, path planning, and traffic control. However, existing models for vehicle trajectory prediction predominantly focus on limited driving scenarios, resulting in limited applicability. To address this limitation, we present a novel vehicle trajectory prediction approach that employs a Convolutional Long Short-Term Memory (CNN-LSTM) model, incorporating simulated environments and vehicle dynamic time series data, including longitudinal, vertical, and latitudinal position and acceleration. Our approach is distinguished by its ability to handle diverse urban driving scenarios, such as highways, roundabouts, intersections, and turns, which enhances its applicability and generalizability. We experimented and collected vehicle data from 17 drivers using a stationary driving simulator and the Euro Truck Simulator software. For the model implementation and validation, we utilized Python 3.9 and Google Colab, as well as the Scikit-learn library for Deep learning algorithms. The proposed CNN-LSTM model leverages a convolutional layer to learn local patterns and an LSTM layer to capture long-term temporal dependencies, improving performance in predicting vehicle trajectories. The experimental results demonstrate that the CNN-LSTM model provides more accurate predictions for longitudinal and lateral positions compared to traditional vehicle trajectory prediction methods that employ LSTM and Recurrent Neural Network (RNN). This research contributes to developing robust and reliable vehicle trajectory prediction systems vital for ADAS and AVs' safe and efficient operation. The proposed approach broadens the applicability of trajectory prediction models, enabling better-informed decision-making in various driving conditions and ultimately improving road safety and efficiency in the rapidly evolving field of autonomous transportation.

The issue of cognitive distraction during driving has been identified as a major cause of road accidents. Detecting cognitive distraction in real-time can be a valuable strategy for preventing accidents. In this study, a novel approach is presented for the purpose of detecting cognitive distraction in real-time using artificial intelligence while taking into account eye-tracking and head movement data, combined with driving performance measures. This methodology involved collecting data from participants in a driving simulator, on a motion platform, while they performed a cognitive task as well as a control driving scenario. The data collected included eye-tracking data, head movement data, driving performance measures, and subjective ratings of distraction. To develop an accurate model for detecting cognitive distraction, a Deep Feedforward Neural Network (D-FFNN) model was employed while considering binocular gaze direction, pupil diameter, orientation of each eye, head rotational velocities, and head acceleration. The developed model was trained using the collected data and achieved an accuracy of 96.09% in detecting cognitive distraction. The results of our study demonstrate the effectiveness of the proposed method in identifying cognitive distraction in real-time. Also, the accuracy of this model was compared with other AI based classification algorithms. The proposed method has significant implications for preventing vehicle accidents caused by cognitive distraction. The proposed method can be integrated into existing driver-assistance systems to alert drivers and assist them in returning their focus to the road.

Advances in utilisng motion simulators for skill training and related applications have yielded numerous benefits, such as safety, availability, and serviceability, environmentally friendly, and economically beneficial. To give simulator users a sense of realistic feeling of driving, an accurate motion cueing algorithm (MCA) is essential, in order to respect the simulator platform limitation and avoid motion sickness. The use of Model Predictive Control (MPC) in MCA designs leads to respecting the constraints and considering the future dynamic behaviors of the simulator. However, the tuning process of the MPC prediction horizon and control horizon still need to be improved. These horizons are normally selected manually by the designer. Previous studies on meta-heuristic algorithms produce a large prediction horizon with a heavy computational load or a small prediction horizon that sacrifices the stability and accuracy of the simulator system. In this study, the Grasshopper Optimization Algorithm (GOA) is adopted to yield optimal prediction and control horizons in MPC-based MCA models. The results are compared with those from the Butterfly Optimization Algorithm (BOA) and Genetic Algorithm (GA) in terms of sensation error and computation time. The GOA technique depicts the fastest process time to promptly detect proper MPC horizons. It does not affect the simulator's efficiency in utilising the workspace, as evidenced by the correlation coefficient and root mean square error between sensation from a real-world vehicle and the simulator.

Driving simulators are utilized for many applications including basic driver training, human factor studies, human-machine interaction, and vehicle prototyping in automobile industries. The main purpose of using driving simulator is to provide realistic driving experience. Since simulator platforms have physical limitations, Motion Cueing Algorithms (MCAs) are used to generate driving sensation for the simulator user while considering the simulator's physical and dynamical constraints. When using a model predictive control (MPC)-based MCA, the principle of MPC is leveraged to predict an optimized future behavior of the simulator where a series of control actions is developed across a defined future horizon using the explicitly specified process model. Corresponding to the pre-positioning or time-varying reference MPC, it is crucial to predict the future vehicle motion signals for the simulator accurately. The existing methods for predicting vehicle motion signals do not excel in predicting time-series of a long sequence due to the missing feedback loop or limited memory size. To address this issue, the Long Short-Term Memory (LSTM) model is developed to predict motion signals using Python. The performance of LSTM is compared with those from different traditional methods using several measurements criteria, which include the root mean squared error (RMSE), mean absolute error (MAE), and Pearson’s correlation coefficient (r). The results indicate that LSTM outperforms RNN by producing more accurate motion allowing the MCA to deliver realistic motion sensations, the LSTM model can be employed in a wide range of applications including autonomous vehicles trajectory prediction, and other prediction problems.

The motion signals are generated for a simulator user based on the visual understanding of the environment using virtual reality. In this respect, a motion cueing algorithm (MCA) is employed to reproduce the motion signals based on the real driving motion scenarios. Advanced MCAs are required to predict precise driving motion scenarios. Nonetheless, investigations on effective methods for predicting the driving motion scenarios accurately are limited. Current state-of-the-art studies mainly focus on the averaged motion signals from several simulator users pertaining to a specific map or from feedforward neural network and non-linear autoregressive. The existing methods are unable to yield precise predictions of the driving scenarios. In this research, the echo state network and long short-term memory models are employed for the first time in MCA to forecast the driving motion signals. Our evaluation proves the efficiency of our proposed methods in comparison with existing methods.

The conflicts between the perceived sensation of the different sensory systems can cause adverse effects which is known as motion sickness (MS) and the side effects of MS include nausea, dizziness, stomach awareness etc. Virtual reality sickness (also called Cybersickness or visually induced motion sickness (VIMS)) happens during exposure to a virtual environment when senses transfer conflicting sensation signals to the brain. The symptoms of Cybersickness are similar to motion sickness symptoms. The adverse effects of this common phenomenon can negatively affect the training outcome and benefits using VR, undermine users’ health and usefulness of simulators as it involves health risk and contributes to the increase of dropout rates. Therefore, to mitigate these issues, MS should be detected and measured. The primary objective of this study is to subjectively and objectively detect and quantify cybersickness level using a helicopter simulator. This study has also investigated the change in cybersickness self-reported scores in different weather conditions such as clear and stormy. Simulator sickness questionnaire (SSQ) has been employed for subjective scoring. This research also aimed to correlate SSQ scores with physiological data such as Galvanic Skin Response (GSR). The findings demonstrated that the SSQ total score (TS) has increased significantly from clear weather to stormy for the participants. There is also a positive correlation found between the change in TS and the amount of GSR but not significant.

People perform most of their activities while being in an upright position. Nonetheless, there are some circumstances where they are required to adapt to different positions other than the upright position as in air manoeuvres and sport gymnastics. In these unexpected scenarios, the physiological signals are likely to change which can affect their performance. While some studies investigated the correlation between physiological signals and different body positions, to our best knowledge, these studies were limited to a rotating chair (1 or 2 degree of freedoms). Here, we investigated and evaluated human physiological responses (such as pupil diameter, skin temperature, heart rate, and breathing rate) to different seated positions including seated supine, seated side, seated inverted, and seated upright using Universal Motion Simulator (UMS), a 6 degree of freedom simulator with the most realistic acceleration and motion sensation. Open loop acrobatic flight motion sensation for the 11 participants were created and accompanied with a series of pre- and post-questionnaires to subjectively assess the physical wellbeing of each participant. The results of the study based on the objective assessment of collected physiological data showed that the mean heart rate decreases during an inverted position (82 beats per minutes bpm) and increased by an average of 7 beats per minute in an upright position 89.5 bpm. Moreover, the mean breathing rate in an upright position was 18.3 respirations per minutes (rpm) which is higher than mean breathing rate in the side position 19.8 rpm. Furthermore, it was found that the mean pupil diameter (PD) in an upright position was 4.33 mm which is higher compared to other positions. Independent from the motion scenarios and body positions, the Skin Temperature kept increasing which might be because of excitement and other emotional factors.