Fecha de publicación: 04/01/2023 Fuente: WIPO (eseential oils OR extracts) An autonomous vehicle (AV 1308) located in a road environment 1400 stores a plurality of vehicle operation constraints. It receives data (1352,Fig.13) from multiple onboard sensors (1344,Fig.13), the data respectively describing the environment and a physical characteristic of a passenger of the AV. From the constraints and sensor data, an onboard processor extracts a feature vector comprising first feature describing an object located within the environment (e.g. another vehicle 1316) and a second feature describing a physical characteristic of the passenger. A machine learning circuit of the AV generates a motion segment 1404 based on the feature vector, such that a number of violations of constraints is below a threshold. The AV autonomously traverses to a destination in accordance with the motion segment, which is either a trajectory between two spatiotemporal locations or a speed that avoids collision between the AV and the object. The sensor data may comprise the passenger’s heart rate, temperature, pupil dilation, facial expression or skin conductance or a pressure applied to a seat arm-rest.