Field of the invention
[0001] This invention relates to the field of providing driver assistance to a driver of a vehicle and more specifically in the field of providing driver assistance based on health of the driver.
Background of the invention
[0002] During driving, many medical conditions can negatively affect the driving. Hence, it is necessary to identify health condition of a driver. According to Chinese patent application, CN104207791, a driving method for detecting fatigue is disclosed. In CN104207791, a method for detecting fatigue driving, belonging to the field of automotive safe driving, by taking behavior index of the driver (steering wheel main data, driving speed and driving time), pre-information, to obtain the fatigue detecting physiological parameters as indicator variables and BP (Back propogation) neural network was trained using a fatigue indicator corresponding variable establishment fatigue detection network model, using the model for fatigue detection.
[0003] However, in the Chinese application only one condition, that is fatigue detection is addressed.
Brief description of the accompanying drawing
[0004] Different modes of the invention are disclosed in detail in the description and illustrated in the accompanying drawing:
[0005] Figure 1 illustrates a flowchart describing a method of providing driver assistance, in accordance with one embodiment of the disclosure.

Detailed description of the embodiments
[0006] A method of providing driver assistance is disclosed. The method comprises receiving (105), by a processor, a plurality of input parameters from a vehicle, the input parameters comprising at least one of a first dataset comprising real-time driving pattern associated with a driver of said vehicle, a second dataset comprising health record of the driver and a third dataset comprising real-time road condition of the vehicle. The method also includes retrieving (110), by the processor, a set of trained-data, the trained-data being stored in a memory unit and comprising an aggregation of a plurality of driving patterns classified into one of a corresponding plurality of medical-conditions and a corresponding plurality of road conditions, the classification being done based on one of a supervised learning technique and an unsupervised learning technique. Also, the method includes comparing (115), by the processor, the input parameters with the set of trained-data to determine a match between the real¬time driving pattern and at least one driving pattern of the plurality of driving patterns present in the set of trained-data and transmitting (120), by the processor, an alert information to at least one of the vehicle and a plurality of vehicles surrounding the vehicle if the match is determined between the real-time driving pattern and at least one driving pattern of the plurality of driving patterns.
[0007] The method is performed by a processor. In one embodiment, the processor is embedded in a server device present in a cloud network. An electronic control unit of a vehicle can connect to the server device present in the cloud network. Also, each vehicle is adapted to communicate with other vehicle through V2V communication or through the cloud network.

[0008] At step 105, the processor receives a plurality of input parameters from a vehicle. An electronic control unit in the vehicle connects to the processor in the cloud network through wireless communication interface. The input parameters include, but are not limited to, a first dataset including real-time driving pattern associated with a driver of the vehicle, a second dataset including health record of the driver and a third dataset including real-time road condition of the vehicle
[0009] The real-time driving pattern comprises data associated to acceleration, deceleration, braking force and turning radius of the vehicle. Various sensors present in the vehicle is used for retrieving this data. The data retrieved from various sensor are transmitted to the processor by the electronic control unit in the vehicle. Health condition of the driver influences the driving pattern. For example, if the driver has high blood pressure then the driving pattern may include sharp decelerations due to sudden braking. Similarly various other medical conditions are associated with a corresponding driving pattern. Therefore, it is necessary to tap the real-time driving pattern. Thus the real-time driving pattern is transmitted to the processor.
[0010] One another input parameter includes a health record of the driver. In one embodiment, the health record is derived from an insurance database. That is, the processor may have access to the insurance database of a specific driver and hence the medical condition of the driver can be determined. In one embodiment, the health record may be obtained using a health-input provided by the driver itself. The health-input can be provided by entering the medical condition on the dashboard which is transmitted to the processor or a voice input can be provided describing the health condition that will be transmitted to the processor present in the cloud network.
[0011] Further, another input parameter includes real-time road condition of the vehicle. The real-time road condition indicates the environment where the vehicle is

being driven. Examples of the real-time road condition include, but are not limited to, a highway, a city-drive, an icy road, an uphill drive and a terrain-drive. The real-time road condition are also transmitted to the processor so that the driving patterns associated with each of these road conditions can be retrieved from a memory unit connected to the processor.
[0012] At step 110, the processor retrieves a set of trained-data. The set of trained-data is being stored in a memory unit that is accessible by the processor. The set of trained-data includes aggregation of a plurality of driving patterns classified into one of a corresponding plurality of medical-conditions and a corresponding plurality of road conditions. Examples of the medical-conditions include, but are not limited to, conditions that affect alertness and memory, learning and judgement, high blood pressure, physical disabilities, cardiovascular disease, neurological conditions, mental disorders, diabetes and vision problems. In other words, the set of trained-data includes driving pattern associated with each medical condition. For example, if the medical condition is high blood pressure then this condition will include an associated driving pattern. Similarly, if the medical condition includes vision problem then it will include an associated driving pattern. Therefore, every medical condition and its associated driving pattern forms a part of the set of trained-data.
[0013] Similarly, every road condition is associated with a corresponding driving pattern that is stored in the memory unit and can be accessed by the processor. For example, an icy road will have a specific driving pattern. Similarly, a highway will be associated with a specific driving pattern. Hence, the driving patterns are classified based on various road conditions which forms a part of the set of trained-data.
[0014] The classification of numerous driving pattern into various medical-conditions or road conditions is performed using supervised learning technique or an

unsupervised learning technique. In other words, driving pattern from numerous drivers with various medical conditions are obtained. Further, driving pattern of drivers having a similar medical condition are grouped using the supervised learning technique or an unsupervised learning technique. Hence, the more driving patterns are analysed, the more accurate will be the grouping. Similarly, the driving patterns from numerous drivers driving in various road conditions are obtained and are grouped. Hence, the more driving pattern data is obtained, the easier it will be for classifying the driving pattern into a specific medical condition or a specific road condition or both. It should be noted that, supervised learning can be used to detect patterns corresponding to a medical condition and unsupervised learning can be used to detect anomalies in the driving patterns.
[0015] Therefore, the set of trained-data is obtained from driving patterns of numerous drivers. Further, supervised learning technique or an unsupervised learning technique is applied on this data (driving patterns of numerous drivers) to classify the driving patterns into various medical-conditions or various road conditions.
[0016] Such trained-data is used as a reference or a model for classifying the real-time driving pattern associated with the driver to any one medical-condition or to any one road conditions. Such reference can be used for predicting the health of driver in real¬time thereby avoiding impending catastrophes.
[0017] At step 115, the processor compares the input parameters with the set of trained-data to determine a match between the real-time driving pattern and at least one driving pattern of the plurality of driving patterns present in the set of trained-data. A comparator present in the processor may be used for such comparison. If there is a match then the corresponding medical-condition associated with the matching data of the trained-data is retrieved by the processor. The matching of the real-time driving

pattern with one of the driving pattern in the set of trained-data indicates that the user is suffering from the medical-condition associated with this driving pattern in the set of trained-data.
[0018] Further if the real-time driving pattern do not match with any of the driving pattern in the set of trained-data, then the processor checks the health record of the driver. If the health record of the driver indicates that the driver is healthy then no action is taken by the processor. However, if the health record of the driver indicates that the driver has a certain medical-condition, then this real-time driving pattern is tracked and forms a part of the set of trained-data. Further, in future if a similar driving pattern is seen in another driver then that driver can be grouped under this specific medical-condition. Therefore, the set of trained-data will learn and be updated automatically and continuously thereby increasing accuracy and reliability of the trained-data.
[0019] At step 120, the processor transmits alert information to at least the vehicle or a plurality of vehicles surrounding the vehicle or both if the match between the real¬time driving pattern and at least one driving pattern of the plurality of driving patterns is determined.
[0020] The alert information comprises one of health-caution-data, automatic driving control instructions and precaution-data. The health-caution-data is transmitted by the processor to the driver cautioning him that based on his driving pattern he may be suffering from a certain medical condition and hence must undergo medical check-up to avoid any untoward incident with regard to his health.
[0021] Similarly, the automatic driving control instructions is transmitted by the processor to the vehicle based on the road condition. For example, based on the driving

pattern, if it is determined that the vehicle is on icy road, then the processor transmits automatic driving control instructions such as “reduced speed” or “braking instruction” to the vehicle.
[0022] Further, the precaution-data is transmitted by the processor to the plurality of vehicles surrounding the vehicle. For example, if the vehicle occupant is an aged person suffering from various medical-condition such as cardiovascular conditions then the precaution-data indication “no honk” or “do not overtake” can be transmitted by the processor to the plurality of vehicles surrounding the vehicle carrying the aged person.
[0023] Therefore, the method enables prediction of health of the driver and further cautioning the driver about any medical emergency which is safety critical. Further, the method also supports for alerting the surrounding vehicles based on the health of the driver so that the surrounding vehicles are cautioned. Also, by updating the trained-data, the reliability and accuracy of the prediction is enhanced.
[0024] It should be understood that embodiments explained in the description above are only illustrative and do not limit the scope of this invention in terms of the type of learning techniques or method of determining the driving patterns and technique used for classification. Many such embodiments and other modifications and changes in the embodiment explained in the description are envisaged. The scope of the invention is only limited by the scope of the claims.

1. A method of providing driver assistance, said method comprising:
receiving (105), by a processor, a plurality of input parameters from a vehicle, said input parameters comprising at least one of a first dataset comprising real-time driving pattern associated with a driver of said vehicle, a second dataset comprising health record of said driver and a third dataset comprising real-time road condition of said vehicle;
retrieving (110), by said processor, a set of trained-data, said trained-data being stored in a memory unit and comprising an aggregation of a plurality of driving patterns classified into one of a corresponding plurality of medical-conditions and a corresponding plurality of road conditions, said classification being done based on one of a supervised learning technique and an unsupervised learning technique;
comparing (115), by said processor, said input parameters with said set of trained-data to determine a match between said real-time driving pattern and at least one driving pattern of said plurality of driving patterns present in said set of trained-data; and
transmitting (120), by said processor, an alert information to at least one of said vehicle and a plurality of vehicles surrounding said vehicle if said match is determined between said real-time driving pattern and at least one driving pattern of said plurality of driving patterns.
2. The method as claimed in claim 1, wherein said real-time driving pattern
comprising data associated to acceleration, deceleration, braking force and turning
radius of said vehicle.

3. The method as claimed in claim 1, wherein said health record is derived from an insurance database and a health-input provided by said driver.
4. The method as claimed in claim 1, wherein said plurality of road conditions comprises a highway, a city-drive, an icy road, an uphill drive and a terrain-drive.
5. The method as claimed in claim 1, wherein said alert information comprises one of health-caution-data, automatic driving control instructions and precaution-data.
6. A processor for providing driver assistance by monitoring health of a driver and surrounding-area of a vehicle of said driver, said processor comprising:
receive a plurality of input parameters from a vehicle, said input parameters comprising at least one of a first dataset comprising real-time driving pattern associated with a driver of said vehicle, a second dataset comprising health record of said driver and a third dataset comprising real-time road condition of said vehicle;
retrieve a set of trained-data, said trained-data being stored in a memory unit and comprising an aggregation of a plurality of driving patterns classified into a corresponding plurality of medical-conditions, said classification being done based on one of a supervised learning technique and an unsupervised learning technique;
compare said input parameters with said set of trained-data to determine a match between said real-time driving pattern and at least one driving pattern of said plurality of driving patterns present in said set of trained-data; and
transmit alert information to at least one of said vehicle and a plurality of vehicles around said vehicle if said match is determined between said real-time driving pattern and at least one driving pattern of said plurality of driving patterns.

7. The processor as claimed in claim 6, wherein said real-time driving pattern comprising acceleration, deceleration, braking force and turning radius.
8. The processor as claimed in claim 6, wherein said health record is derived from an insurance database and a health-input provided by said driver.