Driver abnormality sign detection device

A driver abnormality sign detection device detecting an abnormality sign of a driver driving a vehicle, the driver abnormality sign detection device comprising:

• a control circuit configured to detect the abnormality sign of the driver based on visual line information acquired from a visual line detector, driving operation information acquired from a driving operation detector, and traveling environment information acquired from a traveling environment detector, wherein
• the control circuit is configured to
• run a visual line model, using driving operation and traveling environment as inputs, that outputs a predicted visual line of the driver in a healthy state;
• run a driving operation prediction model, using the traveling environment as an input, that outputs a predicted driving operation value of the driver in the healthy state;
• acquire an actual visual line based on the visual line information,
• calculate a visual line abnormality degree which represents a difference between the actual visual line and the prediction visual line;
• acquire an actual driving operation based on the driving operation information,
• calculate a driving operation abnormality degree that represents a difference between the actual driving operation and the prediction driving operation;
• calculate a traveling risk of the vehicle based on the traveling environment information;
• calculate a comprehensive abnormality degree by combining the visual line abnormality degree, the driving operation abnormality degree, and the traveling risk;
• and
• on condition that the comprehensive abnormality degree is a predetermined threshold value or greater, detect the abnormality sign of the driver.

A non-transitory computer readable storage device having computer readable instructions that when executed by circuitry cause the circuitry to:

• acquire a predicted visual line based on traveling environment information of a vehicle,
• acquire an actual visual line of a driver of the vehicle based on visual line information,
• calculate a visual line abnormality degree based on a difference between the actual visual line and the predicted visual line,
• acquire a predicted driving operation based on the traveling environment information,
• acquire an actual driving operation based on driving operation information,
• calculate a driving operation abnormality degree based on a difference between the actual driving operation differs from the predicted driving operation,
• calculate a traveling risk of based on the traveling environment information,
• calculate a comprehensive abnormality degree by combining the visual line abnormality degree, the driving operation abnormality degree, and the traveling risk, and
• on condition that the comprehensive abnormality degree is equal to or greater than a predetermined threshold value, detect an abnormality sign of the driver on condition.