A Case Study on Vestibular Sensations in Driving Simulators.

Area of Science:

• Engineering
• Human-Computer Interaction
• Vehicle Dynamics

Background:

• Motion platforms are integral to simulators, but physical limitations necessitate motion cueing algorithms (MCA).
• Classical washout filters, widely used in MCAs, are known to cause information loss.
• Inadequate motion simulation can lead to unsatisfactory user experiences and compromised training objectives.

Purpose of the Study:

• To evaluate the fidelity of a 6DOF motion platform simulation using real-world vehicle acceleration data.
• To identify specific events where classical washout filters fail to accurately reproduce vehicle dynamics.
• To assess the impact of lost acceleration data on the adequacy of simulator training.

Main Methods:

• Recorded acceleration data from a BMW 325Xi AUT across all degrees of freedom (DOF).
• Inputted recorded data into a mathematical simulation model including washout filters, kinematics, and actuator simulation.
• Qualitatively compared the simulated motion platform output with the recorded real-world input data.

Main Results:

• The simulation accurately reflected the recorded input for most events.
• Three critical acceleration events were identified where significant data was lost during simulation.
• These lost events are considered vital for effective learning in training simulators.

Conclusions:

• Classical washout filters in motion cueing algorithms demonstrate limitations in reproducing critical vehicle accelerations.
• The identified data loss in specific events can significantly hinder the effectiveness of vehicle simulators for training purposes.
• Further research into advanced MCA is needed to improve simulator fidelity and training outcomes.