WO2014120053A1

WO2014120053A1 - Description behavior detection for safe driving and automobile control based on the detection result

Info

Publication number: WO2014120053A1
Application number: PCT/SE2013/050071
Authority: WO
Inventors: Johan Hjelm
Original assignee: Telefonaktiebolaget L M Ericsson (Publ)
Priority date: 2013-01-30
Filing date: 2013-01-30
Publication date: 2014-08-07

Abstract

Behavior detection for safe driving and automobile control based on the detection result are disclosed. More specifically, in an automobile capable of being controlled by outputs obtained via a bus from plural types of integrated sensors which detect a driving condition of the automobile, the following steps are executed. The driving condition of the automobile is monitored, an action is activated, based on action data, which causes the automobile to move for a short period of time. A driver's response to the action is measured, based on the outputs from the plural types of sensors, and the driver's response is evaluated. Then, it is determined based on a result of the evaluation whether or not the driver is within an acceptable range, and the automobile's movement is controlled, based on a result of the determination.

Description

DESCRIPTION

BEHAVIOR DETECTION FOR SAFE DRIVING AND AUTOMOBILE CONTROL BASED ON THE DETECTION RESULT

TECHNICAL FIELD

[0001] This invention relates to driver' s behavior detection and automobile control based on the detection result .

BACKGROUND

[0002] So far, various types of systems to ensure the safety of driving automobiles have been developed. A collision prevention system using radar, anti-skid system, and speed control system are exemplary examples.

[0003] In addition to the above, a car control system in which if it is detected that a driver is inebriated, the ignition system is disabled have also been developed. More specifically, the ignition system is connected to a breathalyzer, unless a driver first goes through the breathalyzer test (i.e. blows into the mouthpiece), the vehicle cannot start. Such a test indirectly contributes to prevention of inebriated driving .

[0004] For example, U.S. Patent Application

Publication No. 2010/0314190 Al (US 2010/0314190 Al) discloses an interlock system for a vehicle which is coupled with an alcohol tester for measuring the alcohol level of a driver for the vehicle, and is controlled to start the vehicle or prevent the vehicle from being started, depending on the measured alcohol level .

[0005] A similar system was also published in U.S.

Patent Application Publication No. 2011/0304465 Al (US 2011/0304465 Al) . According to US 2011/0304465 Al, the system empirically measures a driver' s cognitive and driving skills required for safe driving, and verifies the driver's identity. More specifically, a

breathalyzer is connected with a mobile phone with a Bluetooth® function which can communicate with a

vehicle's disabling system. When a driver blows into a mouthpiece of the breathalyzer, the measurement result is transmitted to the vehicle's disabling system

through the Bluetooth© unit. Then, depending on the result, the vehicle's disabling system determines whether or not to allow the vehicle to operate.

[0006] A different approach for safe driving has been proposed in US 2009/0284361 Al . US 2009/0284361 Al discloses a lane changing detection and warning system which integrates IR sensors including a

transmitter and receiver in a car, and detects changes in diffusely reflected IR light levels measured by the receiver. On a roadway, white lane markers are painted. Thus, when a car crosses over a white lane marker, the IR sensor senses changes in the received IR levels. If such changes are frequent, the system recognizes that the car frequently changes the lane on a roadway, and concludes that the driving of the driver is dangerous, and warns the driver to slow down.

[0007] US 2009/0063201 Al focuses on a difference in a certain driving action such as a brake depression between a sober driver, an impaired driver, and a driver using a cellular phone while driving. US

2009/0063201 Al discloses the following two parts: an algorithm installed in the car computer system, which passively detects the behavior of an individual and relates it to a known set of behaviors, for generating a report that describes the extent to which the driver has engaged in non-desirable behavior (e.g. drunk driving) ; and a reporting system.

[0008] The algorithm is assumed to be an NILM

(Non-Intrusive Load Monitor) algorithm and is installed in the computer of the car. Although the NILM

algorithm is an algorithm initially developed to

identify and detect electric appliances, US

2009/0063201 Al applies it to impaired driving behavior. In such an application, the algorithm needs to be given a value which is the "normal" value indicating that a driver is not incapacitated and which has to be

collected from somewhere. Usually, this value is collected from academic research in which drivers were asked to operate a driving simulator while legally intoxicated. Then, this is compared with a control group of non-intoxicated drivers. The same measurement is probably made for both drivers operating a cellular phone and drivers not operating a cellular phone. For example, Strayer, D.L. et al. of the University of Utah have made intensive research in finding such a value.

[0009] However, the above-mentioned conventional arts have the following drawbacks in view of enhancing safety driving.

[0010] Current solutions to detect drunk driving include direct measurements on the bodily functions of the driver. These are frequently perceived as

intrusive, and have to be tuned to the inebriation method. In a case where a breathalyzer is used, it is quite easy to deceive the breathalyzer. The

breathalyzer connected to the ignition system of a car can easily be fooled, for instance by a passenger blowing into the mouthpiece. It also does not detect some drugs which can impair driving performance in the same way as alcohol. For instance, meters tuned to detect alcohol inebriation will not detect heroin inebriation .

[0011] In a case where a lane changing detection unit, it is costly since a special IR sensor and recognition unit are required. In the cases in the art where a cognitive test is used for detecting a driver's impairment, it usually takes an extended period of time to complete the test. US 2009/0063201 Al assumes that the period of test is equivalent to the entire driving session of the driver, and the report is assumed to summarize the driving session.

[0012] The detection of non-desirable behaviors by the use of the NILM algorithm is also wholly passive.

[0013] As can be understood from the above, there is no active detection of undesirable behaviors, as have been demonstrated by the tests at the University of Utah, it is still possible for a driver to drive while using a cellular phone or being intoxicated and appear to drive normally until a crisis occurs. Hence, the probability that a dangerous driver is detected is small until an accident is about to happen.

[0014] According to US 2009/063201 Al, the

research study by Strayer, D.L. et al. shows quite different results for impaired drivers and drivers operating a cellular phone. This indicates that it is quite difficult to find an appropriate threshold value applicable to identifying unusual driver' s behavior due to drinking alcohol and/or operating a cellular phone.

[0015] While current solutions to detect and counteract drunk driving work well as long as the driver uses the alcohol tester, no similar measure exists to determine whether or not a driver is using a cellular phone, even though this can cause similar problems in driving. Provided that an attempt to implement countermeasures to keying into the cellular phone while driving a car is made, the implementation has to be based on a network or interception depending on law enforcement. However, such implementation implies an additional cost for the operator of the network, or is sporadic at best.

[0016] Also, current methods of the detection and disablement of vehicles apply only to the use of cars and similar vehicles. Motorcycles, where drivers are equally dangerous not in the least to themselves, do not normally have a technology as the described

installed.

[0017] In addition, operation of a vehicle when a driver is sleep-deprived will more or less create the same effect in the driving pattern of the vehicle, but will not generate any specific metabolites which can be detected by a breathalyzer tune to the detection of alcohol or drug consumption.

SUMMARY

[0018] Accordingly, the present invention is conceived as a response to the above-described

disadvantages of the conventional art.

[0019] The present invention is based on an idea in which actions and other activities will be combined with a number of features, including adaptive learning, to create a mechanism by which a driver's profile can be determined from the way which the driver interacts with various activities in a manner imperceptible to the driver.

[0020] One embodiment of the present invention includes a module, modifying the control program of the vehicle and interacting with it, configured to insert a number of tests or trials into the operation of the vehicle. If the driver does not respond satisfactorily, the normal vehicle operation can be interrupted, for instance slowing it down to a safe speed. Ultimately, after a sufficient number of tests have failed, the vehicle can be stopped.

[0021] To be effective, such tests have to be integrated into the normal traffic situation in a very cunning and subtle manner so that an inebriated or otherwise incapacitated driver cannot be prepared for an artificial situation and cannot respond to it in a way that would be adequate for a sober, but clumsy, driver .

[0022] More specifically, according to one aspect of the present invention, there is provided an

automobile capable of being controlled by outputs obtained via a bus from plural types of integrated sensors which detect a driving condition of the automobile .

[0023] The automobile comprises: a monitor unit configured to monitor the driving condition of the automobile, a memory unit configured to store action data which causes the automobile to move for a short period of time; an activation unit configured to activate an action based on the action data stored in the memory unit; a measurement unit configured to measure a driver' s response to the action, based on the outputs from the plural types of sensors; an evaluation unit configured to evaluate the driver's response; a determination unit configured to determine whether or not the driver' s state is within an acceptable range, based on a result of the evaluation; a control unit configured to control the automobile's movement, based on a result of the determination.

[0024] According to another aspect of the present invention, there is provided a module installed in an automobile capable of being controlled by outputs obtained via a bus from plural types of integrated sensors which detect a driving condition of the

automobile .

[0025] The module may comprise all units in the above automobile other than the control unit.

[0026] According to still another aspect of the present invention, there is provided a method applied to an automobile capable of being controlled by outputs obtained via a bus from plural types of integrated sensors which detect a driving condition of the automobile . [0027] The method comprises the steps of:

monitoring the driving condition of the automobile;

activating an action based on action data, which causes the automobile to move for a short period of time, stored in a memory unit; measuring a driver's response to the action, based on the outputs from the plural types of sensors; evaluating the driver's response;

determining whether or not the driver's response falls outside an acceptable range of responses, based on a result of the evaluation; and controlling the

automobile's movement, based on a result of the

determination .

[0028] According to still another aspect of the present invention, there is provided a computer program to be executed in a computer of an automobile capable of being controlled by outputs obtained via a bus from plural types of integrated sensors which detect a driving condition of the automobile. The computer program comprises the above steps recited in the method.

[0029] According to still another aspect of the present invention, there is provided a computer

readable memory which stores the above computer program.

[0030] This invention makes it possible for an automobile to positively detect a driver's behavior which might cause a traffic accident, and take an appropriate countermeasure for safe driving.

[0031] This contributes to reducing a number of accidents caused by, for example, an inebriated driver or a driver using a cellular phone while driving.

[0032] Other features and advantages of the present invention will be apparent from the following description taken in conjunction with the accompanying drawings, in which like reference characters designate the same or similar parts throughout the figures thereof .

BRIEF DESCRIPTION OF DRAWINGS

[0033] The accompanying drawings, which are incorporated in and constitute a part of the

specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0034] Fig. 1 is a schematic top view of a car integrating various types of sensors, according to an exemplary embodiment of the present invention.

[0035] Fig. 2 is a schematic block diagram showing a detailed construction of a controller in the car.

[0036] Fig. 3 is a functional block diagram showing a detailed construction for checking a driver' s response .

[0037] Fig. 4 is an example schematically showing an unexpected action when a driver starts a car.

[0038] Fig. 5 is an example schematically showing an unexpected action when a driver is driving a car. [0039] Fig. 6 is a conceptual flow chart showing a process for detecting unusual driver's behavior.

[0040] Fig. 7 is a detailed flow chart showing one of typical actions for detecting unusual driver' s behavior when a driver starts a car.

[0041] Fig. 8 is a detailed flow chart showing one of typical actions for detecting unusual driver' s behavior when a driver is driving a car.

DETAILED DESCRIPTION

[0042] An exemplary embodiment of the present invention will now be described in detail in accordance with the accompanying drawings.

[0043] Fig. 1 is a schematic top view of a car integrating various types of sensors, according to an exemplary embodiment of the present invention.

[0044] As shown in Fig. 1, a car 100 integrates various types of sensors for safe driving and automatic control. These sensors are connected to a CAN-bus 210. The CAN-bus 210 is connected to a controller (CNTL) 200 which includes a CPU 201. All outputs from the sensors are transferred to the controller 200, and in response to the transferred information, the CPU 201 issues an instruction to the associated unit or module for controlling the car.

[0045] These sensors includes a front radar 221, rear radar 222, side radars 223-224 for detecting an object in front of the car 100, behind the car 100, and at both sides of the car 100. Wheel sensors 251-254 are provided near respective wheels 101-104 or

respective disc brakes 111-114. A three-dimensional gyro (GYRO) 231, acceleration sensor (SNS) 232, speed sensor (SNS) 233, acceleration pedal sensor 234, brake sensor 235, and steering sensor 241 are also provided for the safe driving. The steering sensor is linked to a steering wheel 141, and the acceleration pedal sensor 234 and brake sensor 235 are respectively linked to depression of respective pedals 142-143.

[0046] An engine 131 is monitored by an engine rotation sensor 242, actuated by an engine actuator 255 under control from the controller 200, and ignited by a driver through an ignition key 145. A driving force from the engine 131 is transferred to the wheels 101- 104 through wheel shafts 121, 122 and an engine shaft 123.

[0047] The CAN-bus is a standard for the

interconnection of car components. Different

manufacturers use different command sets on top of the standardized physical bus. The commands used in this embodiment must be translated to the commands

appropriate for the car where this embodiment is implemented.

[0048] Fig. 2 is a schematic block diagram showing a detailed construction of the controller (CNTL) 200 in the car 100.

[0049] As shown in Fig. 2, the controller 200 is connected to the CAN-bus 210 through an input port 204 and an output port 205. All sensor signals come to the input port 204, and then are transferred to the CPU 201 or an ASIC 203. The ASIC 203 mainly handles the sensor signals required for real-time processing and/or fast response. The CPU 201 executes programs stored in a memory (MEM) 202. The CPU 201 and the ASIC output control signals and/or instructions to relevant units {not shown) through the output port 205.

[0050] For example, when the front radar 221 detects an unexpected obstacle in front while driving a car, the detected signal is immediately transferred to the ASIC 203. In response to the radar detection signal, the ASIC 203 real-timely issues an instruction to the engine actuator 255 or a braking system (not shown) so that the car can immediately slow down or stop.

[0051] As summarized, the car 100 which integrates various types of sensors is controlled by the

controller 200 based on the sensor signals.

[0052] It goes without saying that a modern car integrates a number of sensors other than sensors mentioned above.

[0053] For example, the above four (4) radars 221-

224 may be replaced by four active optical sensors such as laser radars (lidars) , respectively, to detect an object or obstacles in the respective directions.

[0054] Behavior detection for safe driving and automobile control based on the detection result in the car described above will now be explained with

reference to the drawings.

[0055] In this embodiment, the controller 200 occasionally creates various types of actions which require a driver's quick response, and actively work on the car 100 according to the created action. Then, the controller 200 monitors a driver's response based on signals outputted from various sensors, determines whether or not the driver's response falls within an acceptable range of responses, and controls the car to stop or slow down if it is determined that the driver' s response to the action is not within the acceptable range of responses.

[0056] Fig. 3 is a functional block diagram showing a detailed construction for checking a driver' s response .

[0057] In this embodiment, a module 270 in the

ASIC 203 plays a main role in creating various types of actions which require a driver's quick response. The module 270 includes a monitor unit 270a, a check unit 270b, a selection unit 270c, an activation unit 270d, a measurement unit 270e, an evaluation unit 270f, and a determination unit 270g. [0058] As shown in Fig. 3, the module 270 is connected to the CAN-bus 210 via an interface module 260, a range database 271, rule database 272, action database 273, and history database 274. The interface module 260 is connected to a command database 275.

[0059] The (reasoning) module 270 described here executes a set of rules stored in the rule database 272, based on inputs received from sensors and the like, and values stored in the range database 271 and previously executed tests and records stored in the history

database 274. For example, when the input signal indicating that the car 100 is started comes from the ignition key 145, the rule is to compose a list of tests. Further, when the input values (e.g. brake pedal depression) received from the sensors (e.g. brake sensor 235) connected to the CAN-bus 210 diverge more than a certain amount (a pre-determined range) stored in the range database 271 from a preset value, an action can be triggered, such as initiating a second test or shutting down the vehicle.

[0060] When the module 270 executes a set of rules, and satisfies a certain condition (e.g. when a driver starts the car, or when the car is triggered by some other condition (e.g. a passive monitoring such as the NILM monitoring described in US 2009/0063201 Al ) ) , the module 270 then initiates to execute a certain action stored in the action database 273. The other conditions for triggering the car to do something are not further discussed here, but can also include the triggering remotely by law enforcement. This raises the security issue, but could also be included among the initiation methods.

[0061] To enable the driver to realize whether a cognitive test was performed or the car has

malfunctioned, an indicator such as a display or instrument panel light can be turned on as the test is initiated, to indicate that a test was performed.

[0062] The policies encoding the actions stored in the action database 273 and executed by the module 270 can be expressed in a suitable language, e.g. the FIPA Agent Communication Language, which is used for the communication between intelligent agents such as the controllers of traffic lights and controllers of autonomous vehicles.

[0063] Example policies are (when expressed in human language) :

"Slow down and veer left". This can be done in a way which simulates e.g. side wind, but only for a few seconds. If the driver is reacting normally, the compensation for the simulated side wind will take place within a few tenths of a second and the driver will not overcompensate . On the other hand, if the driver is drunk, the reaction is likely to be more sluggish and the driver is likely to overcompensate. The same reaction will also be sluggish if the user is talking on a cellular phone while driving a car, but the overcompensation is likely less.

"Increase acceleration by 2 km per hour for 2 seconds, if driving speed is faster than 20 km per hour". This is likely to lead a sober driver to step on the brakes instantly, whereas an impaired driver may wait. But, a speed increase of 4 km per hour is not likely to be a security risk (even when the increase is from 20 to 24 km, the driver should be able to

compensate without endangering anyone) .

[0064] The actions can be specifically selected to determine whether a driver is in an impaired state, based on research such as that reported from the

University of Utah.

[0065] As exemplified above, the action database

273 stores a set of actions to be executed. It is noted that actual commands to the car according to the executed actions are issued from the interface module 275 to pertinent units in the car through the CAN-bus 210, while referring to the command database 275 which stores a set of instructions/commands.

[0066] In another possible implementation, this embodiment can be implemented as a set of triggers and stored procedures in a database system. Embedded database systems have now become small enough yet powerful enough to implement real-time or near-real- time applications.

[0067] It is obvious for those skilled in the art that this invention can naturally also be implemented in several other ways, for instance as a real-time system based on the sensor readings from the CAN-bus.

[0068] Figs. 4 and 5 respectively illustrate executed actions in the car 100 when the car is about to start and when the driver is driving the car, based on scenarios which may occur during the course of driving .

[0069] Fig. 4 depicts one typical example of an action to be executed when a driver starts a car. In Fig. 4, the left side shows that the car is not moving, while the right side shows that the car has just started. In the right side figure, the broken line indicates an actually accelerated state while the solid line indicates an expected accelerated state.

[0070] More specifically, when the driver turns the ignition key 145 and presses on the acceleration pedal 142, the module 270 execute an action which causes the car 100 to move at a speed slightly faster than an expected speed by the depression of the acceleration pedal 142 for a few seconds. Then, the module 270 monitors the deceleration of the car in response to the unexpected acceleration.

[0071] This action is performed when the front radar 221 confirms that nobody is in front of the car 100 for safety purpose.

[0072] Fig. 5 depicts one typical example of actions to be executed when a driver is driving a car. In Fig. 5, the lower side shows that the car is running in a tunnel 500, while the upper side shows that the car has just come out of the tunnel and has just come across side wind. In the upper side figure, the broken line indicates an actually steered state while the solid line indicates an expected steered state.

[0073] More specifically, when the driver comes across the side wind immediately after the tunnel from the right side with respect to the moving direction, and steers the steering wheel 141 rightward, the module 270 execute an action which causes the car 100 to steer in a direction slightly more leftward than an expected direction for a few second, for example. Then, the module 270 monitors the moving direction of the car in response to the unexpected movement.

[0074] This action is performed when the front, rear and side radars 221-224 confirm that there are sufficient distances to a side guard rail, cars in front, and cars behind for safety purpose.

[0075] However, the action illustrated in Fig. 5 is not necessarily performed when the car has actually come out of the tunnel. For example, the action may be performed when the car is coming out of the parking lot.

[0076] Fig. 6 is a conceptual flow chart showing a process for detecting unusual driver's behavior. In this process, we presume that the above-described database such as the Range Database (DB) 271, Rule Database (DB) 272, Action Database (DB) 273, History Database (DB) 274, and Command Database (DB) 275 are provided in the controller 200.

[0077] At step S100, the module 270, particularly the monitor unit 270a, monitors whether or not a condition based on a set of rules stored in the Rule DB 272 and/or a value based on a set of range values stored in the Range DB 271 is satisfied for triggering a certain action. If the condition is satisfied, the process advances to step S110. On the other hand, the condition is not satisfied, the module 270 continues to monitor .

[0078] At step S110, the module 270, particularly the check unit 270b, checks the History DB 274 which stores the previous records as to what action has been taken during the previous few driving sessions. At step S120, based on the check result, the module 270, particularly the selection unit 270c, selects currently available actions from the Action DB 273 so that a driver will not experience the same test frequently.

[0079] At step S130, the module 270, particularly the activation unit 270d, creates a list of potentially available actions, randomly selects one of the actions from the list, and sends information on the selected action to the interface module 260. In response to the information, the interface module 260 accesses the

Command DB 275, and initiates the action by

transmitting necessary commands or instructions to a pertinent unit via the CAN-bus 210. The commands or instructions cause the car 100 to somewhat move in an unexpected way for a driver.

[0080] At step S140, the module 270, particularly the measurement unit 270e, waits for the driver's response, and measures the driver's response to the action through the sensor outputs, from the control mechanisms of the car (the steering wheel, brakes etc) , all of which are fed through the CAN-bus 210.

[0081] At step 150, the module 270, particularly the evaluation unit 270f, evaluates the driver's

response, and at step S160, the module 270 examines whether or not the drive's state is clear. If the driver's state is unclear, the process returns to step S130 to select another action from the list. In the examination, the module 270 compares the driver's response with parameters which relate to driver' s behavior such as impaired driver's behavior.

[0082] On the other hand, if the driver's state is clear, the process advances to step S170. At step S170, the module 270, particularly the determination unit 270g, determines whether or not the drive's state is normal. If the driver's behavior is normal (i.e. within the acceptable ranges as stored in the range database 271 and the rule database 272), the process advances to step S180 to continue driving. After this, the process returns to step S100 to monitor whether or not the above condition is satisfied.

[0083] However, if the driver' s behavior is abnormal (falling outside the acceptable ranges) indicating e.g. impaired driving, the process advances to step S190. At step S190, the module 270 requests the CPU 201 to stop the car or slow down the car. In this case, as an alternative to stopping the car completely, it may be possible to operate at reduced effect, e.g. driving with a maximum speed of 30 km per hour .

[0084] After step S190, law enforcement can optionally be alerted, provided that the vehicle is equipped with an appropriate communications system.

[0085] As explained above, the module 270

according to this embodiment provides: (1) a mechanism which causes the car 100 to provide a driver' s test according to commands and instructions when the car is activated; and (2) a mechanism which receives the driver's response and control the car's operation according to the driver's response.

[0086] Next, two specific examples will now be described with reference to Figs. 7-8.

[0087] Fig. 7 is a detailed flow chart showing one of typical actions for detecting unusual driver's behavior when a driver starts a car.

[0088] At step S200, the module 270 confirms whether or not the ignition key 145 is turned by a driver. If the ignition key is turned, the process advances to step S210 to check whether or not there is any one in front of the car 100 by the front radar 221. In a case where somebody is in front of the car, the process advances to step S 280 to pick up another action, and then advances to step S290 to execute the other test according to the picked action.

[0089] At step 210, in a case where it is

confirmed that nobody is in front of the car, the process advances to step S220. At step S220, the module 270 issues via the interface module 260 an instruction by which the car is moved at a speed slightly faster than an expected speed for the driver for a few seconds when the driver depresses the

acceleration pedal 142. Then, at step S230, the module 270 measures a response time until the driver depresses the brake pedal 143 and the depression force. The response time and the depression force are obtained from outputs of the brake sensor 235.

[0090] At step S240, the module 270 evaluates the driver's response based on the measured response time and the measured depression force. The depression profile over time may be included for the evaluation. [0091] At step S250, the module 270 determines, based on the evaluation result, whether or not the driver's state falls within the acceptable range. If it is determined that the driver's state is acceptable, the process advances to step S260 to allow the driver to continue driving. After this, the module 270

continually monitors the driver's state.

[0092] On the other hand, if it is determined that the driver's state is not acceptable (i.e. the response falls outside the acceptable range) , the process

advances to step S270 at which the module 270 requests the CPU 201 to stop the car or slow down the car.

[0093] Fig. 8 is a detailed flow chart showing one of typical actions for detecting unusual driver' s behavior when a driver is driving a car.

[0094] At step S300, the module 270 confirms whether or not the car 100 is running in the tunnel 500 by using radar signals from the front, rear and side radars 221-224. If it is confirmed that the car has just come out from the tunnel 500, the process advances to step S310 to check whether or not there is any car in front of the car 100 by the front radar 221, any car behind the car 100 by the rear radar 222, and any car at both sides of the car by the side radars 223-224. In a case where another car is nearby and there is not a sufficient distance from the car 100 to the other car, the process advances to step S380 to pick up another action, and then advances to step S390 to execute the other test according to the picked action.

[0095] At step 310, in a case where it is

confirmed that no car is nearby or there is a

sufficient distance from the car 100 to the other car, the process advances to step S320. At step S320, the module 270 issues via the interface module 260 an instruction by which the car is steered in a direction slightly more sideway than a direction where the driver intended to steer for a few seconds when the driver steered the steering wheel 141. Then, at step S230, the module 270 measures a response time until the driver steers back and a steering angle. The response time and the steering angle are obtained from outputs of the steering sensor 241.

[0096] At step S340, the module 270 evaluates the driver's response based on the measured response time and the measured angle. The angle profile over time may be included for the evaluation.

[0097] At step S350, the module 270 determines, based on the evaluation result, whether or not the driver's state is normal. If it is determined that the driver's state is normal, the process advances to step S360 to allow the driver to continue driving. After this, the module 270 continually monitors the driver's state .

[0098] On the other hand, if it is determined that the driver's state is abnormal, the process advances to step S370 at which the module 270 requests the CPU 201 to stop the car or slow down the car.

[0099] Two (2) specific examples have been

described above. However, it goes without saying that another action (test) such as simulating wheel spin on an icy road may be created. To do this, for safety purpose, it may possible to simply increase a number of rotations in the engine 131 without connecting the driving force to the wheel shafts 121-122 so that the car 100 does not move but the driver becomes aware of the sudden increase of the number of rotations of the engine 131.

[0100] It should be noted that although the module

270 is integrated in the ASIC 203 in this embodiment, all processes described above may be implemented in the CPU 201 by executing a real-time software program which is stored in the memory 202.

[0101] It should also be understood that the module 270 does not exactly detect that a driver is impaired or a driver is using a cellular phone while driving. Rather, the module 270 occasionally attempts to detect the driver's response while driving by actively causing the car to somewhat move in a manner unexpected by the driver. This endangers the driver to invest mental effort in driving a car, reducing the risk of other types of distracted driving (e.g. noticing a friend of the sidewalk) as well.

[0102] It also contributes to reducing the

personal tragedy for the drunk driver.

[0103] According to the embodiment as described above, the following other beneficiary effects are achieved :

(1) It reduces the damages to road infrastructure caused by collisions caused by impaired drivers;

(2) It raises the value of a car where the above embodiment is implemented, since it reduces a number of accidents in a case where a car is carelessly driven; and

(3) It serves as a reminder for cellular phone users to switch off their cellular phones, if the above embodiment of this invention is executed.

[0104] As mentioned above, the above described driver's behavior detection can be implemented by a software program which a computer in a car executes.

If a faster response or processing is required, such detection and car control may be implemented in an ASIC. Therefore, a software program is part of this invention.

[0105] As many apparently widely different

embodiments of the present invention can be made

without departing from the spirit and scope thereof, it is to be understood that the invention is not limited to the specific embodiments thereof except as defined in the appended claims.

Claims

1. An automobile (100) capable of being controlled by outputs obtained via a bus (210) from plural types of integrated sensors which detect a driving condition of the automobile, comprising:

a monitor unit (270a) configured to monitor the driving condition of the automobile;

a memory unit (273) configured to store action data which causes the automobile to move for a short period of time;

an activation unit (270d, 260) configured to activate an action based on the action data stored in the memory unit;

a measurement unit (270e) configured to measure a driver's response to the action, based on the outputs from the plural types of sensors;

an evaluation unit (270f) configured to evaluate the driver's response;

a determination unit (270g) configured to determine whether or not the driver's state is within an acceptable range, based on a result of the

evaluation; and

a control unit (200, 201) configured to control the automobile' s movement, based on a result of the determination .

2. A module (270) installed in an automobile (100) capable of being controlled by outputs obtained via a bus (210) from plural types of integrated sensors which detect a driving condition of the automobile, the module comprising:

an evaluation unit {21 Of) configured to evaluate the driver's response; and

a determination unit (270g) configured to determine whether or not the driver' s is within an acceptable range, based on a result of the evaluation.

3. The module according to claim 2, further

comprising :

a check unit (270b) configured to check what action has been taken during a previous few driving sessions; and

a selection unit (270c) configured to select currently available action data from the memory unit, wherein the selection unit excludes action data which has been used for performing an action during the previous few driving session from the selection, and the memory unit stores history information as to what action has been taken during the previous few driving sessions.

4. The module according to claim 3, wherein the activation unit activates an action based on the action data selected by the selection unit.

5. The module according to any one of claims 2 to 4, wherein the activation unit issues a command or

instruction via the bus to pertinent portions in the automobile causing the automobile to move according to the command or instruction.

6. The module according to any one of claims 2 to 5, wherein the determination unit is further configured to request a CPU (201) in the automobile to stop or slow down the automobile, based on a result of the

determination .

7. The module according to any one of claims 2 to 6, wherein the bus includes a CAN-bus.

8. The module according to claim 7, wherein the memory unit stores a set of rules (272) and a set of values (271) defining the driving condition of the automobile .

9. The module according to claim 8, wherein the plural types of sensors includes:

a front detection sensor (221) for detecting an object in front of the automobile;

a rear detection sensor (222) for detecting an object behind the automobile;

side detection sensors (223,224) for detecting an object at both sides of the automobile; and

a steering sensor (241) for detecting a steering angle actuated by a steering wheel (141); and

a brake sensor (235) for detecting a depression force of a brake pedal (143) and a depression profile over time.

10. The module according to claim 9, wherein when a driver starts the automobile, if it is confirmed by at least part of the plural types of sensors that nobody is in front of the automobile, the activation unit is further configured to activate an action data which causes the automobile to move.

11. The module according to claim 10, wherein if it is confirmed by the monitor unit, based on the set of rules that an ignition key (145) for the automobile is turned,

the check unit is further configured to check whether or not any one is in front of the automobile, using detection data from the front detection sensor, and

if nobody is in front of the automobile, the activation unit is further configured to activate an action data which causes the automobile to move at a speed slightly faster than an expected speed by a depression of an acceleration pedal (142) for a few second .

12. The module according to claim 11, wherein the measurement unit is further configured to measure a response time until a driver depresses the brake pedal (143) and the depression force, based on outputs of the brake sensor (235).

13. The module according to claim 12, wherein the evaluation unit is further configured to evaluate the driver' s response based on the measured response time and the measured depression force.

14. The module according to claim 9,

wherein if it is confirmed by the monitor unit, based on the set of rules and detection data obtained from the side detection sensors (223, 224) that the automobile has come out of a tunnel (500),

the check unit is further configured to check whether or not there is any other automobile or

obstacle nearby the automobile using detection data from the front detection sensor, the rear detection sensor and the side detection sensors,

if there is neither automobile nor obstacle nearby the automobile, the activation unit is further configured to activate an action data which causes the automobile to steer in a direction slightly more sideway than a direction where a driver intended to steer by the steering wheel (141) for a few second.

15. The module according to claim 14, wherein the measurement unit is further configured to measure a response time until the driver steers back and a steering angle, based on outputs of the steering sensor (241) .

16. The module according to claim 15, wherein the evaluation unit is further configured to evaluate the driver's response based on the measured response time and the measured angle.

17. The module according to claim 3, wherein the monitor unit, the activation unit, the measurement unit, the evaluation unit, the determination unit, the check unit and the selection unit are integrated into an ASIC (203) .

18. A method applied to an automobile (100) capable of being controlled by outputs obtained via a bus (210) from plural types of integrated sensors which detect a driving condition of the automobile, comprising the steps of:

monitoring (S100) the driving condition of the automobile;

activating (S130) an action based on action data, which causes the automobile to move for a short period of time, stored in a memory unit (273);

measuring (S140) a driver's response to the action, based on the outputs from the plural types of sensors;

evaluating (S150) the driver's response;

determining (S170) whether or not the driver's state is within an acceptable range, based on a result of the evaluation; and

controlling (S180, S190) the automobile's

movement, based on a result of the determination.

19. A computer program to be executed in a computer of an automobile capable of being controlled by outputs obtained via a bus (210) from plural types of

integrated sensors which detect a driving condition of the automobile, comprising the steps of:

monitoring (S100) the driving condition of the automobile;

evaluating (S150) the driver's response;

controlling {S180, S190) the automobile's movement, based on a result of the determination.

20. A computer readable memory which stores a computer program according to claim 19.