Grounding E-Scooter Motion Cueing Parameters in Rider Biomechanics; Simulated Trust: Tools, Rules, and the Mutual Shaping of Autonomous Vehicle Safety 2 views

Both projects in this portfolio address the same problem: simulation outputs treated as valid representations of physical conditions without demonstrated evidence that they are. When that mapping is left unexamined, simulation-validated systems fail on deployment, simulators destabilize the humans who use them, and regulatory frameworks inherit the blind spots of the tools they oversee. The technical report revisits a virtual reality e-scooter simulator whose 3DOF motion platform produced rider-reported instability during simulated grade changes. The cause was not tuning but two architectural defects in the controller: a heave gravity-subtraction error saturating the vertical axis on every frame, and unfiltered terrain pitch reaching the platform faster than a standing rider can react to. Grounding three cueing limits in rider biomechanics and platform geometry rather than designer preference, the revised washout filter eliminates the gravity error and bounds tilt onset to the rate a rider can follow. The STS paper asks the same question at the institutional scale: has trust in autonomous vehicle simulation been earned or assumed? Analysis of federal policy, the technical literature, and seven major autonomous driving developers between 2016 and 2026 finds that no regulatory standard defines when synthetic evidence becomes admissible, that federal policy presupposes a technical standardization the industry does not exhibit, and that the October 2023 Cruise dragging incident illustrates the cost of leaving simulation fidelity unexamined. Together, the projects argue that simulation is trustworthy only when its mapping to physical reality is demonstrated rather than assumed.

BS (Bachelor of Science)

Virtual Reality; e-scooter; motion-cueing; mutual shaping

English

2026-05-11