System and method for managing driver takeovers of autonomous vehicles based on monitored driver behavior

Abstract

A method of managing operator takeovers of an autonomous vehicle is provided. The method includes collecting information about an external environment of the autonomous vehicle; analyzing the collected information about the external environment of the autonomous vehicle to determine an upcoming traffic pattern; collecting information about an operator of the autonomous vehicle; analyzing the collected information about the operator of the autonomous vehicle to determine operator behavior; predicting operator action based on the determined upcoming traffic pattern and the determined operator behavior; and initiating a predetermined vehicle response based on the predicted operator action. Predicting operator action includes comparing the determined upcoming traffic pattern to similar historical traffic patterns and retrieving historical operator actions in response to the similar historical patterns.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to vehicles equipped with advanced driver assistance systems, and more particularly, to a system and method of managing driver takeover of advanced driver assistance systems based on a monitoring behavior of a vehicle operator. BACKGROUND

[0002] Advanced Driver Assistance Systems (ADAS) are intelligent systems located on a vehicle that assist a driver (also referred to as a vehicle operator) in the operation of the vehicle. ADAS are used to enhance or automate selective motor vehicle systems in order to improve the driving performance of the vehicle operator or to improve the level of autonomous driving in accordance with SAE J3016 levels of driving automation. A typical ADAS includes an ADAS module that communicates with various vehicle exterior sensors, vehicle state sensors, and selective vehicle control systems such as steering, acceleration, and braking systems. The ADAS module analyzes information collected by the exterior sensors and vehicle state sensors to generate instructions and communicate them to the vehicle control systems, thereby providing partial or full automatic control of the vehicle. The ADAS can also include a Driver Monitoring System (DMS) having a DMS module that communicates with various vehicle interior sensors configured to monitor the behavior of the vehicle operator, e.g., eye gaze, facial expressions, body movements, and other factors related to the subject, to predict the fatigue, distraction, and emotional state of the vehicle operator.

[0003] In one operational scenario, when the ADAS is operating in a lower level of autonomous driving mode (i.e., SAE J3016 levels 0-2) and the DMS detects that the vehicle operator can be fatigued or distracted, the DMS can initiate an audible or visual alert to warn the vehicle operator and / or communicate with the ADAS module to take over control of the vehicle from the vehicle operator. In another operational scenario, when the ADAS is operating in a higher level of autonomous driving mode (i.e., SAE J3016 levels 3-5) and the ADAS module encounters a driving scenario that can require manual control of the vehicle, the ADAS can instruct the DMS to activate an audible or visual alert to request that the vehicle operator manually control the vehicle. In yet another operational scenario, when the ADAS is operating in a partial to full autonomous driving mode and the vehicle operator does not have sufficient confidence that the ADAS can adequately handle the traffic conditions, the vehicle operator can voluntarily take over control of the ADAS.

[0004] The vehicle operator taking over control of the ADAS is referred to as takeover control of the autonomous vehicle, or simply takeover. The ADAS requesting the vehicle operator to manually control the vehicle is referred to as handover control of the autonomous vehicle, or simply handover.

[0005] Accordingly, while vehicles equipped with ADAS having DMS achieve their intended purpose, there is a continuing need for improvement to improve the quality of experience of the vehicle operator by reducing the perceived need or desire for the vehicle operator to take over control from the ADAS and by reducing the frequency of handover requests by the vehicle operator from the ADAS. SUMMARY

[0006] A method of managing operator takeovers of an autonomous vehicle is provided. The method includes collecting information about an external environment of the autonomous vehicle, analyzing the collected information about the external environment of the autonomous vehicle to determine an upcoming traffic pattern, collecting information about an operator of the autonomous vehicle, analyzing the collected information about the operator of the autonomous vehicle to determine operator behavior, predicting operator action based on the determined upcoming traffic pattern and the determined operator behavior, and initiating a predetermined vehicle response based on the predicted operator action. Predicting operator action includes comparing the determined upcoming traffic pattern to similar historical traffic patterns and retrieving historical operator actions in response to the similar historical patterns.

[0007] A method of managing the intent of a vehicle operator to take over control of an autonomous vehicle due to a perceived need or desire is disclosed. The method includes collecting information about an upcoming traffic pattern by at least one external sensor, collecting information about behavior of the vehicle operator by at least one internal sensor, analyzing the behavior of the vehicle operator in response to the upcoming traffic pattern to determine when the vehicle operator perceives a need to take over control of the autonomous vehicle, and initiating a dynamic change in the autonomous vehicle to eliminate the perceived need for the vehicle operator to take over control of the autonomous vehicle.

[0008] A method of managing warning priority of a vehicle operator is provided. The method includes collecting external information about an environment surrounding the vehicle, collecting internal information about the vehicle operator, analyzing the external information to determine an upcoming traffic pattern, analyzing the internal information to determine operator behavior in response to the upcoming traffic pattern, predicting operator action based on the determined operator behavior in response to the determined upcoming traffic pattern, and prioritizing warnings based on the predicted operator action.

[0009] Further areas of applicability will become apparent from the description provided herein. It should be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure. Attached Figure Description

[0010] The accompanying drawings described herein are for illustrative purposes only and are not intended to limit the scope of this disclosure in any way.

[0011] Figure 1 This is a functional diagram of an autonomous vehicle equipped with an advanced driver assistance system (ADAS) according to an exemplary embodiment, the autonomous vehicle having a driver monitoring system (DMS);

[0012] Figure 2 Management according to exemplary embodiments Figure 1 A block diagram of a method for driver takeover in an autonomous vehicle;

[0013] Figure 3 This is a functional diagram of an autonomous vehicle system according to an exemplary embodiment;

[0014] Figure 4 It is a plan view of a traffic pattern according to an exemplary embodiment, which may cause a perceived need or expectation of a vehicle operator to take over control of an autonomous vehicle;

[0015] Figure 5 This is a flowchart illustrating a method for managing perceived needs or desires to take over control of an autonomous vehicle according to an exemplary embodiment.

[0016] Figure 6 A plan view of a traffic mode according to an exemplary embodiment is shown, which can cause the ADAS to issue a warning priority to the vehicle operator in order to take over control of the autonomous vehicle.

[0017] Figure 7 This is a flowchart of a method 700 for changing the warning priority of a vehicle operator according to an exemplary embodiment; and

[0018] Figure 8 It is a plan view of a traffic pattern according to an exemplary embodiment, wherein driver scanning behavior can be used to predict takeover. Detailed Implementation

[0019] The following description is exemplary in nature and is not intended to limit the scope, application, or purpose of this disclosure. Illustrated embodiments are disclosed with reference to the accompanying drawings, in which the same numbers denote corresponding parts in several figures. The drawings are not necessarily drawn to scale, and some features may be enlarged or reduced to show detail of particular features. The specific structural and functional details disclosed are not intended to be construed as limiting, but rather as a representative basis for teaching those skilled in the art how to practice the disclosed concepts.

[0020] As used herein, a module or control module refers to any one or various combinations of one or more processors, related memory, and other components that are operable under software control to provide the function of the module. Processors include, but are not limited to, dedicated, embedded, distributed, distributed-memory, and shared-memory machines. Related memory includes, but is not limited to, read-only memory (ROM), random access memory (RAM), and electrically programmable read only memory (EPROM). The functions of the control modules set forth in the disclosure can be implemented in a distributed control architecture, where several networked control modules are employed to provide the functionality of the control modules. Control modules can include various communication interfaces, including point-to-point or discrete wiring to other control modules, as well as wired or wireless interfaces.

[0021] Software, firmware, program, instructions, routines, code, algorithms, and similar terms mean any control module executable set of instructions including methods, calibrations, data structures, and lookup tables. Control modules have a set of control routines executed to provide the described functionality. Routines are executed, for example, by a central processing unit, and are operable to monitor inputs from sensing devices and other networked control modules, and are operable to execute control and diagnostic routines to control operation of actuators. Routines can be executed periodically during ongoing vehicle operation. Alternatively, routines can be executed in response to occurrence of an event, software call, or according to a requirement input or requested via a user interface.

[0022] Figure 1 A functional diagram of an exemplary vehicle 100 equipped with an advanced driver assistance system (ADAS) 102 is shown, the vehicle 100 having a driver monitoring system (DMS) 104. The ADAS 102 is configured to provide a driving automation level from a partially automated driving mode to a fully automated driving mode in accordance with SAE J3016 driving automation levels. Lower driving automation levels can include a range of dynamic driving and vehicle operation including some level of automatic control or intervention related to simultaneous automatic control of multiple vehicle functions (e.g., steering, acceleration, and braking) with the operator retaining partial control of the vehicle. Higher driving automation levels can also include full automatic control of all vehicle driving functions including steering, acceleration, braking, and performing maneuvers such as automatic lane changes, with the driver relinquishing most or all control of the vehicle for a period of time. The vehicle 102 is also referred to as an autonomous vehicle 102.

[0023] The ADAS 102 includes an ADAS module 106, also referred to as an ADAS control module 106, which is configured to communicate with various systems of the vehicle 100, e.g., a detection system 128, an acceleration system 130, a steering system 132, a navigation system 136, a positioning system 138, a deceleration system 140, and other systems necessary to partially or fully control the motion, speed, direction, etc. of the vehicle 102. The DMS 104 includes a DMS module 108, also referred to as a DMS control module 108, which is configured to communicate with the ADAS module 106 and receive data from at least one internal sensor 150 configured to monitor a vehicle operator (not shown). These vehicle systems 128, 130, 132, 136, 138, 140 can have system-specific control modules (not shown) that communicate with the ADAS module 106 for coordinated control of the vehicle 102. In alternative embodiments, the ADAS module 106 can function as a master control module for directly controlling all system-specific control modules or working in conjunction with the system-specific control modules to control one or more of the systems 128, 130, 132, 136, 138, 140.

[0024] The detection system 128 communicates with external sensors 152, including but not limited to optical laser devices, e.g., Light Detection and Ranging (LIDAR) devices 152A for 360° observation of the host vehicle 102, forward-looking cameras 152B, rearward-looking cameras 152C, side-looking cameras 152D, and distance sensors 152E, e.g., radar and sonar devices. The detection system 128 communicates with internal sensors 150, including but not limited to cameras. Each of these internal sensors 150 and external sensors 152 can be equipped with local processing components that process collected data and provide processed or raw sensor data directly to one or more of the detection system 128, the ADAS module 106, and the DMS module 108.

[0025] The vehicle 102 can also include a communication system 142 having circuitry configured with a dedicated short-range communication protocol, e.g., WiFi, for communicating with other vehicles equipped with similar communication systems. The communication system can be configured for vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), and vehicle-to-everything (V2X) communication.

[0026] Communication between ADAS 102, DMS 104, vehicle systems 128, 130, 132, 136, 138, 140, 142, internal sensors 150, and external sensors 152 can be implemented using a direct wired point-to-point link, a networked communication bus link, a wireless link, or another suitable communication link 170. Communication includes the exchange of data signals in a suitable form, including, for example, electrical signals via a conductive medium, electromagnetic signals via air, optical signals via an optical waveguide, and the like. Data signals can include discrete, analog, or digitized analog signals representing inputs from sensors, actuator commands, and communication between vehicle systems and modules.

[0027] Figure 2 A functional block diagram of a method 200 of managing takeovers of an automated vehicle 102 by a driver is shown. The method 200 improves the quality of experience of an operator of an automated vehicle by (i) reducing or eliminating the perceived need or desire of the operator for a takeover action, and (ii) reducing or eliminating the frequency or number of alerts, warnings, or notifications generated for the operator to initiate a takeover action.

[0028] An operator of an automated vehicle is also referred to as an operator of the vehicle, a vehicle operator, an operator, or simply a driver. A takeover action or takeover is defined as an action by the vehicle operator to initiate a takeover of the functionality of the ADAS, also referred to as a takeover of the automated vehicle. Examples of a takeover action include, but are not limited to, the vehicle operator taking over operational control of the vehicle from the ADAS by inputting commands to a steering device, depressing an accelerator pedal, and / or depressing a brake pedal. The intent or motivation of the vehicle operator to initiate a takeover can be due to a perceived need or desire of the vehicle operator to take over control.

[0029] In block 202, external sensors 152 collect information of the external environment of the vehicle 102. In block 204, the communication system 142 can wirelessly receive information about the external environment of the vehicle 102 from a roadside unit or other vehicles equipped with V2V or V2X communication. The information collected from the external sensors 152 and wireless communication can include surrounding vehicle layout, vehicle dynamics, road geometry, weather, lighting conditions, and other necessary information for the ADAS modules to perceive and negotiate through the upcoming traffic pattern. Examples of the upcoming traffic pattern include, but are not limited to, the layout or geometry of the road in the path of the automated vehicle, vehicles traveling on the road, objects on the road that the automated vehicle will need to negotiate through or bypass, and environmental context such as weather and lighting conditions.

[0030] Moving from block 202 and block 204 to block 206, the information collected by the external sensors 152 and V2X communication is analyzed to determine the upcoming traffic pattern.

[0031] In block 208, the internal sensors 150 collect information about the operator of the autonomous vehicle 102. The information collected by the internal sensors 150 is analyzed to determine the behavior of the operator. The facial expression, eye gaze, body posture including gestures, and other subject related factors of the operator are analyzed in blocks 210, 212, 214, and 216, respectively. The subject related factors include fatigue, situational awareness, trust, etc.

[0032] Moving to block 222, the DMS predicts the operator action based on the determined operator behavior and the determined upcoming traffic pattern. The DMS module 108 performs the prediction model by retrieving historical data from block 218. The historical data includes a plurality of historical traffic patterns and a plurality of historical operator behaviors and resulting actions corresponding to the plurality of historical traffic patterns. The DMS module 108 compares the determined upcoming traffic pattern to similar historical traffic patterns and retrieves the historical operator actions corresponding to the similar historical patterns. The DMS module 108 then predicts the trend and probability of the operator’s take-over action by comparing the determined upcoming traffic pattern and observed operator behavior to the historical traffic patterns and historical operator behavior.

[0033] Each vehicle operator has its own personalized prediction model based on its specific historical data. Each newly determined upcoming traffic pattern and corresponding determined operator behavior can be added to the historical data. Referring back to block 222, based on the new and historical traffic patterns and the operator behavior from block 218 in response to these new and historical traffic patterns, an optimization algorithm can be utilized to more accurately predict the trend and probability of the operator take-over. The optimization algorithm can be stored in and executed by the DMS module 108.

[0034] An example operator behavior that can be used to predict operator take-over and change vehicle dynamics can be the operator’s saccadic behavior. The eye gaze of the operator can be analyzed to determine the region of interest, fixation duration, rapid saccade amplitude, etc.

[0035] In block 220, the external information collected in blocks 202 and 204 is communicated to the ADAS module. In block 220, the predicted trend of take-over / no take-over and probability of take-over action of the operator from block 222 is also communicated to the ADAS module. The ADAS module communicates with the vehicle system modules 300 to perform a change in vehicle dynamics or vehicle maneuver to either eliminate the perceived need for a take-over action or preemptively alert the operator of a take-over.

[0036] Reference Figure 3From block 220, the ADAS module can communicate with a steering control module 302 for controlling the EPS motor; an acceleration control module 306 for controlling an engine control module (ECM) 308, a torque control module (TCM) 310, and a power inverted module (Power Inverted Module TCM) 312; and an electronic brake control module (EBCM) 316 for controlling the brakes 318. In block 320, the priority of the system activation escalation can be predetermined based on the severity of the upcoming traffic pattern and the observed operator behavior.

[0037] Example 1 - modifying vehicle dynamics

[0038] Figure 4 A plan view 400 of a traffic pattern that can cause a perceived need for a vehicle operator to take over control of an autonomous vehicle is shown. Figure 5 A block diagram showing a method 500 of managing a perceived need for a vehicle operator to take over an autonomous vehicle by modifying the dynamics of the autonomous vehicle is shown.

[0039] Reference is made to Figure 4 , the plan view 400 shows a two-lane road 401 divided into a first lane 402 and a second lane 404 by a longitudinal dashed line 406. A third lane 408 is shown merging with the first lane 402. An autonomous vehicle 410 is shown traveling in the first lane 402 at a time stamp 1 (Tl) before the third lane 408 intersects the first lane 402. A target vehicle 412 is shown traveling in the third lane 408 toward the first lane 402 at a time stamp 1 (Tl).

[0040] With reference to Figure 4 and Figure 5 , the method 500 begins at block 502 when the DMS analyzes information collected by the external sensors 152 and determines that the upcoming traffic pattern is an upcoming traffic pattern as shown in Figure 4 , the target vehicle 412 merges into the first lane 402 in which the autonomous vehicle 410 is traveling.

[0041] Moving to block 504, the DMS analyzes information collected by the internal sensors and determines that the operator is not paying attention to the road in response to Figure 4the upcoming traffic pattern. This behavior can be interpreted as the vehicle operator having low confidence in the ADAS negotiating this traffic pattern or the vehicle operator being concerned about a potential collision with the target vehicle 412. The behavior of the vehicle operator can be determined based on the vehicle operator's exhibited eye gaze, facial expression, body posture, and other relevant biometrics.

[0042] Moving to block 506, the DMS predicts the potential takeover of the vehicle operator based on historical data about the historical behavior of the vehicle operator in similar historical traffic patterns as Figure 4 the upcoming traffic pattern. The DMS predicts the probability of the vehicle operator taking over the autonomous vehicle to avoid merging with the target vehicle 412. The DMS can also communicate with the ADAS to determine the probability of a collision with the target vehicle 412 to merge.

[0043] Moving to block 508, if the probability of the vehicle operator taking over control of the autonomous vehicle is above a predetermined takeover level or if the probability of a collision with the target vehicle 412 to merge is above a predetermined collision level, the method 500 proceeds to block 510. The term "or" is defined as "inclusive or," meaning either or both. In block 510, the DMS communicates with the ADAS to modify the dynamics of the autonomous vehicle 410 in block 510 to turn the autonomous vehicle at the location indicated by the time stamp 2 (T2) which is later than the time stamp 1 (T1) to the second lane 404 (e.g., change lanes) when the target vehicle 412 to merge approaches the location indicated by the time stamp 2 (T2).

[0044] Moving from block 510 to block 512, if the vehicle operator initiates a takeover, this means that changing the vehicle dynamics from block 510 to mitigate a collision is not the way the vehicle operator wants it. In block 504, the data point is recorded in the historical data. If the vehicle operator does not initiate a takeover, the method proceeds to block 514 and ends.

[0045] Referring back to block 508, if the probability of the vehicle operator taking over control of the autonomous vehicle is at or below a predetermined takeover level and the probability of a collision with the target vehicle 412 to merge is at or below a predetermined collision level, the DMS does not intervene to manage the operator taking over the autonomous vehicle and ends in block 514, then the method proceeds to block 514 and ends.

[0046] Example 2 - changing warning priority

[0047] Figure 6 A plan view 600 of a traffic pattern is shown that can cause the ADAS to issue a warning or alert to the vehicle operator to take over control of the autonomous vehicle. Figure 7 A block diagram of a method 700 to change the priority of a warning to the vehicle operator is shown.

[0048] Referring Figure 6 , plan view 600 shows a roadway 602 having a first lane 604 and a second lane 606 in the opposite direction separated by a dashed line 608. A first safety island 610 is disposed in roadway 602 separating the two opposing lanes 604, 606 to define a first roundabout 612, and a second safety island 614 is disposed further along roadway 602 separating the two opposing lanes 604, 606 to define a second roundabout 616. An autonomous vehicle 618 is shown traveling in first lane 604, entering first roundabout 612 at time stamp 1 (Tl) and entering second roundabout at a later time stamp 2 (T2).

[0049] Referring Figure 6 and Figure 7 , method 700 begins at block 702 when DMS analyzes information collected by external sensors 152 and determines that the upcoming traffic pattern is a first roundabout, followed possibly by a second roundabout.

[0050] Moving to block 704, DMS analyzes information collected by internal sensor(s) and determines the behavior of the vehicle operator in response to the upcoming traffic pattern. The behavior of the vehicle operator can be determined based on the vehicle operator's exhibited eye gaze, facial expression, body posture, and other relevant biometric factors.

[0051] Moving to block 706, DMS searches a historical database to determine whether the vehicle operator has experienced a similar traffic pattern as shown in Figure 6 If the historical data does not show that the vehicle operator has experienced a similar traffic pattern as shown in Figure 6 , then method 700 moves to block 714 and ends. If the historical data does show that the vehicle operator has experienced a similar traffic pattern as shown in Figure 6 , then method 700 continues to block 708.

[0052] Moving to block 708, DMS analyzes information collected by internal sensors 150 to determine the probability that the vehicle operator will take over the autonomous vehicle. If the determined probability is at or below a predetermined level, then the method moves to block 714 and ends.

[0053] Referring back to block 708, if the determined probability is above a predetermined level, the DMS communicates with the ADAS to cancel the pending handoff escalation, e.g., by not sounding a driver alert. Moving to block 712, if the vehicle operator does not take over the autonomous vehicle, the method moves back to block 704 and continues. A failed handoff also means that the prediction was inaccurate. The data point is then recorded in the historical data in block 704 for future alignment. Referring back to 712, if the vehicle operator does take over the autonomous vehicle, the method moves to block 714 and ends.

[0054] Example 3 - using saccadic behaviour to predict takeover

[0055] Figure 8 A plan view 800 of a traffic pattern is shown in which a vehicle operator's scanning behavior can be used to predict a take-over action and change vehicle behavior to preempt that take-over action. Plan view 800 shows a two-lane road 801 divided by a longitudinal dashed line 806 into a first lane 802 and a second lane 804. An autonomous vehicle 810 is shown traveling in the first lane 802, a first target vehicle 812 is shown in the first lane 802 ahead of the autonomous vehicle 810, and a second target vehicle 814 is shown in the second lane 804 adjacent to the autonomous vehicle 810.

[0056] The autonomous vehicle 810 is shown approaching the first target vehicle 812 at a longitudinal approach speed (V Lo ) and a longitudinal approach distance (D Lo ). When V Lo exceeds a predetermined longitudinal approach speed and / or D Lo is less than a predetermined longitudinal approach distance, and the vehicle operator's eye gaze is on a predetermined target for more than a predetermined time limit, the ADAS can adjust the speed of the autonomous vehicle 810 to increase the relative approach distance between the autonomous vehicle 810 and the first target vehicle 812 to preempt the driver's take-over action. For example, the predetermined longitudinal approach speed can be 5 miles / hour or faster, the predetermined longitudinal approach distance can be 29 meters or less, the predetermined scanning time limit can be 2 seconds or more, and the predetermined scanning target can be the first target vehicle 812. Alternatively, a predetermined operator scanning pattern can be used to predict a driver take-over.

[0057] The second target vehicle 814 is shown approaching the autonomous vehicle 810 at a lateral approach speed (V La ) and a lateral approach distance (D La ). When V La exceeds a predetermined lateral approach speed and / or D LaWhen the lateral proximity distance is less than the predetermined lateral proximity distance and the vehicle operator's gaze is on the predetermined target for more than the predetermined time limit, the ADAS can increase the lateral proximity distance or increase the lateral overlap between the vehicles to preempt the driver's take-over action. For example, the predetermined lateral proximity speed can be 1 mile / hour or faster, the predetermined lateral proximity distance can be 0.7 meters or less, the predetermined glance time limit can be 1 second or more, and the predetermined glance target can be the second target vehicle 814. Alternatively, the predetermined operator glance pattern can be used to predict the driver's take-over.

[0058] The above disclosed systems and methods improve the quality of experience of the vehicle operator by reducing the perceived need or expectation of the vehicle operator to take over control from the ADAS and by reducing the handover request frequency of the vehicle operator to take over control from the autonomous vehicle.

[0059] The description of the present disclosure is merely exemplary in nature and variations that do not depart from the spirit and scope of the present disclosure are intended to be within the scope of the present disclosure. Such variations are not to be regarded as a departure from the spirit and scope of the present disclosure.

Claims

1. A method of managing operator takeovers of an autonomous vehicle, comprising: collecting, by at least one external sensor, information about an external environment of the autonomous vehicle; analyzing the collected information about the external environment of the autonomous vehicle to determine an upcoming traffic pattern; collecting, by at least one internal sensor, information about an operator of the autonomous vehicle; analyzing the collected information about the operator of the autonomous vehicle to determine an operator behavior in response to the upcoming traffic pattern; predicting, based on the determined operator behavior in response to the upcoming traffic pattern, an operator action to determine a probability that the operator is about to take over the autonomous vehicle, wherein the operator action is a takeover action; and based on the predicted operator action, initiating a predetermined vehicle response if the determined probability is above a predetermined level, wherein the predetermined vehicle response comprises one of (i) initiating a vehicle maneuver to eliminate the takeover action, and (ii) eliminating a handoff alert to the operator; if the operator does not take over the autonomous vehicle or the handoff fails, a data point is recorded in historical data for future alignment.

2. The method of claim 1, wherein, The predicting the operator action comprises searching a database containing a plurality of historical traffic patterns and a plurality of historical operator behaviors and resulting actions in response to the plurality of historical traffic patterns.

3. The method of claim 2, wherein, The predicting the operator action further comprises comparing the determined operator behavior in response to the upcoming traffic pattern to similar historical traffic patterns and historical operator behaviors and resulting actions in response to the similar historical traffic patterns.

4. The method of claim 1, wherein, The predetermined vehicle response comprises a vehicle maneuver that eliminates a perceived need for the predicted operator action.

5. The method of claim 1, wherein, The analyzing the collected information about the operator of the autonomous vehicle to determine an operator behavior in response to the upcoming traffic pattern comprises analyzing at least one of eye gaze, facial expression, body movement, and posture of the operator.

6. The method of claim 1, wherein, The collecting and analyzing of information about an external environment of the autonomous vehicle and information about the operator of the autonomous vehicle are performed simultaneously.

7. The method of claim 1, wherein, The collecting information about an external environment of the autonomous vehicle comprises wirelessly receiving external information through vehicle-to-everything (V2X) communication.

Images