Abstract: Aspects of the disclosure relate to analyzing head movements in a test driver tasked with monitoring the driving of a vehicle operating in an autonomous driving mode. For instance, a sensor may be used to capture sensor data of a test driver's head for a period of time. The sensor data may be analyzed to determine whether the test driver's head moved sufficiently enough to suggest that the test driver is engaged in monitoring the driving of the vehicle. Based on the determination of whether the test driver's head moved sufficiently enough, an intervention response may be initiated.

Abstract: Assessing a likelihood of a person experiencing a fatigue event when the person tasked with monitoring a vehicle operating in an autonomous driving mode may include receiving a set of response times for a psychomotor vigilance test administered to the person. The test may include a plurality of trials which involve a person lifting a finger from a user input device. Whether the person passed or failed each trial of the set of trials may be determined. A model trained using data from prior psychomotor vigilance tests administered to the person may be identified for the person. Results of the determinations of whether the person passed or failed each trial of the set of trials may be input into the model in order to determine a value representative of a likelihood of a fatigue event. An intervention response may be initiated based on the value.

Abstract: Assessing a likelihood of a person experiencing a fatigue event when the person tasked with monitoring a vehicle operating in an autonomous driving mode may include receiving a set of response times for a psychomotor vigilance test administered to the person. The test may include a plurality of trials which involve a person lifting a finger from a user input device. Whether the person passed or failed each trial of the set of trials may be determined. A model trained using data from prior psychomotor vigilance tests administered to the person may be identified for the person. Results of the determinations of whether the person passed or failed each trial of the set of trials may be input into the model in order to determine a value representative of a likelihood of a fatigue event. An intervention response may be initiated based on the value.

Abstract: Aspects of the disclosure relate to models for estimating the likelihood of fatigue in test drivers. In some instances, training data including videos of the test drivers while such test drivers are tasked with monitoring driving of a vehicle operating in an autonomous driving mode may be identified. The training data also includes driver drowsiness values generated from one or more human operators observing the videos. The training inputs and outputs may be used to train the model such that when a new video of a first test driver is input into the model, the model will output an estimate of a likelihood of fatigue for that test driver.

Abstract: The disclosure relates collecting feedback from passengers of autonomous vehicles. For instance, that a triggering circumstance for triggering a feedback request has been met may be determined. The triggering circumstance may include a driving event, a presence of other road users, or a trip state. A display requirement and data collection parameters for the feedback request are identified based on the determination. The display requirement defines when the feedback request is displayed and the data collection parameters identify information that the feedback request is to collect. The feedback request is provided for display based on the display requirement and data collection parameters. In response, feedback from a passenger of the autonomous vehicle is received and stored for later use.

Abstract: Aspects of the disclosure relate to models for estimating the likelihood of fatigue in test drivers. In some instances, training data including videos of the test drivers while such test drivers are tasked with monitoring driving of a vehicle operating in an autonomous driving mode may be identified. The training data also includes driver drowsiness values generated from one or more human operators observing the videos. The training inputs and outputs may be used to train the model such that when a new video of a first test driver is input into the model, the model will output an estimate of a likelihood of fatigue for that test driver.

Abstract: Assessing a likelihood of a person experiencing a fatigue event when the person tasked with monitoring a vehicle operating in an autonomous driving mode may include receiving a set of response times for a psychomotor vigilance test administered to the person. The test may include a plurality of trials which involve a person lifting a finger from a user input device. Whether the person passed or failed each trial of the set of trials may be determined. A model trained using data from prior psychomotor vigilance tests administered to the person may be identified for the person. Results of the determinations of whether the person passed or failed each trial of the set of trials may be input into the model in order to determine a value representative of a likelihood of a fatigue event. An intervention response may be initiated based on the value.