Simulateurs de conduite et adéquation architecture / modèles : impact sur le « mal du simulateur »

“Simulator sickness,” which is to some extent similar to motion sickness, is particularly problematic. Understanding the causes of simulator sickness and identifying ways to reduce its occurrence are important issues insofar as, on the one hand, the “at the door” filtering of the population studied on the simulator (people who are not susceptible) introduces bias in the results acquired and, on the other hand, a driver who suffers from simulator sickness is likely to behave differently in terms of their driving, which also introduces bias. The hypothesis we defend in this thesis concerns the relationship between the "controllability" of the system (the simulator) and simulator sickness. In driving simulation, the driver is an integral part of the loop: they initiate the movements and receive the system responses. Our hypothesis is that, in this context, a poor control of the simulated vehicle movement induces erratic movements, perceived by the driver as not corresponding to their expectations (because of their previous driving experience) and therefore unacceptable. To evaluate this hypothesis, we have redesigned the rider/simulator interfacing modalities of a pre- existing motorcycle simulator: hardware and software architecture, sensor/actuator structure of the steering column, and implemented vehicle model. In its current version, the simulator allows a higher degree of controllability by the driver for medium to high speeds.