Field of invention:
The present invention relates to the field of automotive industry and devises means of assessing a driver's credit worthiness which enables him to avail credit or loan facilities with ease.
Background and prior art of the invention:
According to the census in India, in 2011, less than 10% of population had access
to credit. This leaves a large part of the adult population asking for more. The
traditional players such as banks & other financial institutions do not find these
customers to be worthy of the risk based on their earnings and ticket size.
This risk averseness is further cemented by the fact that large percentage of this
population is unbanked or under-banked. India has the 2nd largest unbanked
population in the world with nearly 190 million adults without bank accounts.
While this number has decreased post de-monetization, it is still significant
enough.
There are several types of loans in the market that can be broadly divided into
secured and unsecured loans. In secured loans, a loan is provided against an asset
or collateral and no collateral or security is provided for an unsecured loan.
Three primary factors drive lending - Identity, ability to repay and willingness to
repay. Identity is a factor that is provided by identity proofs such as Aadhaar
card, Voter Identity card and the like.
Ability to pay is part of the credit worthiness evaluation process of the individual
seeking the loan. This includes understanding the income and liabilities on that
income. Liabilities of a person in terms of existing loans are established using a
credit bureau in regular casesbut obtaining bureau information for the unbanked
and undocumented customers remains a challenge.
Traditionally, basic checks required include income proof, credit history,
employment history, repayment capability among others and these are largely
solved by income proofs, bank statements and tax returns. The large segment of

the population which are either unbanked or with undocumented earnings do not
fit the bill for this organized formal banking sector.
The formal banking and finance sector comprises traditional banks and NBFCs
(Non-Banking Financial Companies). Outside this sector, the customer or loan
applicant is only left with the option of private lenders which translates to huge
interests and loan sharks, taxing a customer who does not "appear" credit worthy
to the formal sector.
To provide unsecured loans to this section of customers at a reasonable risk
premium, will require a higher involvement of technology and alternate sources
of data to evaluate the credit worthiness of such a customer. Income proofs may
not be from reliable sources of data and default risk for such sectors is much
higher than average, dissuading traditional banking and financial companies from
entering such a customer/product segment.
Secured loans, or loans against an asset, is a problem that is more readily solvable,
allowing a segment of the customers outside the formal sector to gain access to
credit.
Ability to pay is a key challenge addressed by our invention. Ability to pay
comprises 2 parts:
• Demonstrate credit worthiness - How to demonstrate the credit worthiness of an unbanked and undocumented customer so that s/he becomes eligible for a loan, and
• Reducing default risk of the secured asset - includes traceability and remote control of the underlying secured asset.
The prior art CN101739615A relates to a taxi intelligent comprehensive supervision and service system which that enables in the effective management and supervision of a taxi-service. The system through a wireless network tracks the off and on duty of a driver, passenger volume and other data and helps in the analysis of service data including driver service quality level data, public

information data, customer satisfaction data. However, through the technology, the system only enhances the supervision of the taxi industry and passenger satisfaction and there is no scope for credit assessment of a driver.
US8140359B2 refers to an invention in the field of automobile insurance, more specifically, a system and method of determining an objective driver score that may be used universally by independent insurers to evaluate the risk of insuring a driver by employing certain risk assessment factors concerning the driver's personal history which are each analysed to derive weighted scores. However, this invention too makes no reference to estimating the earning potential of a driver.
The present invention overcomes the challenges posed by the prior art by creating a tool, using a combination of hardware and software, that demonstrates the credit worthiness and to reduce the default risk and hence to help underwrite such customers and encourage more lenders to target this segment of unbanked and undocumented customers with a trackable asset.
Objective of the invention
Accordingly, the objective of the present invention is to create a credit risk assessment framework for providing secured loan against a trackable asset to the automotive sector. The invention aims to record an average data on the periodic earnings or income generated by taxi drivers/commercial vehicle drivers thereby formulating a means to showcase their earning capacity while applying for credit from a financial institution. A trackable asset can include a vehicle or even a vital component such as a battery pack. While the present embodiment of the invention is with respect to the automotive sector and specifically targets e-rickshaw or e-taxi (electric vehicles) industry with lithium ion battery pack as an underlying asset, no part of this claim shall be seen as a limitation in its

application across the automotive sector or to providing a secured loan against any kind of trackable asset.
Summary of the invention
The present invention aims to create a credit risk assessment framework for providing secured loan against a trackable asset to the automotive sector. According to the embodiments of the present invention, a credit risk assessment framework is created by providing the below mentioned data. This data can be used to substitute salary slips, income information etc. which are pivotal sources to ascertain income. The present embodiment of the system provides 4 main outputs categorized as follows:
1) Credit worthiness: This provides assessment of the income and the potential liabilities due to operational risk for the driver.
2) Daily earnings - This is estimated from the distance travelled and the load carried. This data is used to calculate the income generated based on kilometres travelled and detecting the presence of extra load other than the driver to indicate that he is carrying a passenger. As the primary application of this invention is for the rickshaw/taxi market, the ability to identify the presence of a passenger is of key importance.
3) Driving behaviour - The invention creates a risk profile of the driver and hence deduces the probability of getting into accidents and 3rd party payouts. In other words, it gives a better understanding of the probability of 3rd party claims on the income earned.
4) Default risk: Possibilities such as flight risk and non-payment of loans are
addressed through a system which allows for the tracking and remote control
of the asset. As the asset is used to generate primary income for the customer
(driver), this ability to track and control the asset gives a superior edge in
reducing the default risk.

• Trackability of the asset: - The invention reduces the flight risk and the default risk of the customer by providing the real-time GPS location of the asset (and hence the owner), allowing the lender to repossess the asset and dispose it at the market value.
• Remote control (on/off) of the asset: by having this ability to remotely control the asset at all points in time.
These outputs are generated on an on-going basis and are further integrated by banking and NBFCs, into their underwriting model to get a more detailed understanding of the customer and dramatically improve his/her chances of accessing credit at a lower risk premium.
This access to continuous information helps reduce the risk premium for the driver on a continuous basis and this reduced risk premium is passed on to the customer (or driver).
Brief Description of figures
The foregoing summary, as well as the following detailed description of preferred embodiments, is better understood when read in conjunction with the appended figures, however, the invention is not limited to the specific assembly and methods disclosed in the said figures.
The aforesaid figures describe an embodiment of the present invention. The said figures in no way restrict the scope of the present invention as described in the specification and accompanying drawings.
1. Figure 1 describes the process of data communication in the system as a preferred embodiment of the invention
2. Figure 2 describes the possible output data received by the server (6) on various parameters of the assessed driver as one of the embodiments of the instant invention

3. Figure 3 describes excerpt of the data formulated by the server (6) on the income generated by a driver on a single day. This data shall be used by Banks, NBFCs or other financial institutions to ascertain whether the specific driver could be eligible for credit.
Detailed description of the invention
The embodiments herein and the various features and details thereof are explained more fully with reference to the non-limiting embodiments that are illustrated in the accompanying drawings and detailed in the following description. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention which will be limited only by the appended claims. Descriptions of well-known components and processing techniques are omitted so as to not unnecessarily obscure the embodiments herein. The illustrations and examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, they should not be construed as limiting the scope of the embodiments herein.
Input parameters:
The credit risk assessment system comprises of the following parts:
• Load sensor (1) - A load cell converts the pressure or the force being measured into weight. According to a preferred embodiment, strain gauge type or piezo-electric type of load cells are used.
• Accelerometer (2) - Accelerometer sensors will measure both the acceleration and the direction of the vehicle. There are 3-axis and 2-axis accelerometers

which indicates the number of axes measured. For an application such as e-rickshaw or e-taxi, 2-axis accelerometers will suffice. For larger electric vehicles, 3-axis accelerometers would be required.
Sensor (3): The preferred embodiment for the instant invention is a GPS sensor. The GPS module of the GNSS (Global navigation satellite system) is used to periodically obtain the real-time latitude and longitude of the vehicle. Internet of Things (IOT chip) (4) - A preferred embodiment for the instant invention is a typical GSM chip which enables 2G/3G/4G machine-to-machine (M2M) communication by transferring the on-board sensor data to the server (6).
Front camera (5) to provide live traffic feed - On-board camera (5) (facing the road ahead) with cloud storage is a preferred method to obtain a live frontal traffic feed from the camera. Other dashboard cameras have a local storage which cannot be used for live feed
Cloud based server system (6)- In addition to the above input data collection sensors, the overall credit risk assessment system comprises a cloud based server system (6) to which all the input parameters attained by the means of sensors and camera (5) are registered. This is achieved by integrating a GSM module (4) with the data fetched from the sensors to the server. Additionally, the details of the driver such as driver identification, trip data and the driver driving behaviour for the trip are calculated on the server constantly.
Output parameters from the system:
1) Creditworthiness
) Driver Earnings: Daily earnings are calculated based on the distance
(number of kilometres) a driver has driven with a passenger. A simple formula
is used to calculate the driver earnings. Earnings per kilometre is calibrated
according to the prevalent market rates. The load cell (1) and GPS sensors (3)

are sampled every 5 seconds to calculate the earnings only during the presence of a passenger or goods on board. The driver earnings are calculated using the following formula:
Driver earnings/day = Rupees/km * distance travelled (in kms) * passenger load (Yes or No/0 or 1)
Example of Output:

Sample # Latitude, Longitude Load present (0/1)
1 41.5,-99.4 0
2 38.5,-98.3 1
Distance travelled between sample #1 and #2 is calculated using the Haversine formula below. For the above latitude, longitude pair of GPS co-ordinates in sample 1 and 2, the distance travelled as calculated is 347.3km.
The distance (in the present embodiment measured in kilometres) travelled is then plugged into the formula below to calculate the driver's earnings.
Driver earnings/day = Rs.lO/km * 347.3kms travelled * 1 (with passenger) = Rs.3473/day
The drivers daily earnings give an indication of the income potential of the driver (both in terms of amount and periodicity) and hence the ability to repay, thereby giving an insight into his/her inherent credit worthiness.
A load cell (1) is placed in the rear seat of the vehicle to detect the presence of a passenger or load. Any weight indicated by the load cell (1) is calibrated

to deliver a "yes" or "no" or a "1" and "0" for presence and absence of passenger respectively.
Distance travelled is calculated using the standard Haversine formula for finding distance between any two points on a spherical surface. When two sets of latitude and longitude are provided - (latl, longl) and (lat2, long2), then the distance between the two is calculated as below:
Delta lat = lat2 - latl (in radians)
Delta long = long2 - longl (in radians)
A = sin(delta lat/2)A2 + cos (latl in radians) *cos(lat2 in radians) *sin(delta
long/2) A2
C = 2*asin(sqrt(A))
R = radius of earth in kms = 6372.8 kms
Distance travelled = R*C
Example of Output:
Nebraska, USA (Latitudel: 41.507, longitude 1: 99.436) Kansas, USA (Latitude2: 38.504, longitude2: 98.316) Using the above formula, the distance comes to 347.3kms
b) Driving behaviour: Driver driving behaviour is classified into high and low risk. Inputs from on-board camera (5) and speed/accelerometer sensor (2) are used to detect sudden braking and impact & whether they were warranted or not. Accidents and sudden braking or near misses caused by objects in front are detected through a combination of front facing camera (5) and accelerometer (2) which detects impact/shock. Accidents that cause impacts from the back of the vehicle are detected through accelerometer (2). An accelerometer measures the force of impact by the effect on the sensor.

Piezo-electric or piezo-resistive accelerometer can be used as preferred embodiments. Impact is detected by sudden change in acceleration.
Sudden change in acceleration along with distance of object detected in front using a camera (5) indicates a possible impact in front if the object detected using the front camera (5) feed is less than 10m away or just sudden change in acceleration, without any proportionate change in current consumed, denotes that there is an impact possibly at the back caused by another vehicle. Either ways, this event is classified as high risk and when a pre-determined frequency of these high risk events is exceeded in a specified period of time, the driver is classified as high risk else s/he is classified as low risk.
Object detection and distance determination is done using image processing algorithms. Current development in the neural network based CNN (Convolution Neural Network) models, object detection applications are easier to develop than ever before. Other than huge execution enhancements, these strategies have additionally been utilizing huge image datasets. Furthermore, with current methodologies focussing on full end-to-end pipelines, performance has also improved significantly, enabling real-time use cases.
The invention uses multiple models like YOLO (You look only once by Redmon, Joseph, and Farhadi, Ali), Fast R-CNN (by Microsoft research), SSD (Single Shot Multibox Detector by Google) etc. for achieving the best optimal results of object detection. We are able to accurately detect objects in front of vehicles and determine the distance.
Example of output:
For a driver who brakes hard or frequently, risk = High

For a driver who drives smoothly, risk = Low
2) Default Risk
a) Trackability of the asset: This is provided through the on-board GPS sensors (3) that provide the latitude and longitude of the asset, in a real¬time fashion. These points can be converted to physical locations and can be pin pointed on a geographical map using relevant APIs respective map service providers.
b) Remote control of the asset: The trackable asset which is the battery pack in this particular use case can be controlled on/off via cloud server (6). A specific command (to shut down to the battery management system or start it) can be sent via the server (6) to specific asset via the GSM communication chip (4). This command can be used to shut down or turn on a battery pack. This gives complete control over the asset in a remote manner.
Because of the availability of the real-time use & tracking of the asset along with the earning details, this data can be used for improving credit worthiness and decreasing default risk. While this data will not be available for each customer/driver at the time of underwriting the loan, this data will be available on an on-going basis. This insight further reduces the risk premium which can then be passed on to the customer.
All this data helps build a financial profile for the customer (driver) which can be used by any Bank or financial institution to extend credit or other financial products. This data can be accessed by any such player via an API from our server
(6).

The above mentioned outputs can be used in any risk management, or premium pricing module used by Bank or financial institution or any such entity. The above data would also help refine the internal risk scoring models used by these entities.
Having described but a preferred embodiment of the invention, it will be appreciated that variations can be made thereto without departing from the spirit or scope of the invention. A number of implementations have been described as examples. Nevertheless, other implementations exist and are covered by the following claims.

We claim,
1. A system for assessing a driver's credit worthiness comprising of the integration of Load Sensor (1), accelerometer (2), sensors (3), IOT module (4) , On board camera (5), and a cloud server system (6) to record an average data on the periodic earnings or income generated by taxi drivers/commercial vehicle drivers and formulating a means to showcase their earning capacity while applying for credit/loans from a financial institution/ or any other entity.
2. A method to record an average data on the periodic earnings or income generated by taxi drivers/commercial vehicle drivers and formulating a means to showcase their earning capacity while applying for credit/loans from a financial institution/ or any other entity, comprising the steps of:

(a) Retrieving information from the Load Sensor (1), accelerometer (2), sensors (3), IOT module (4) and On board camera (5) and registering the respective parameters on to the Cloud Server system,
(b) Calculating the distance travelled, via the GPS co-ordinates retrieved from the sensors (3) and by further using the Haversine formula.
(c) Distinguishing the distance travelled by the vehicle with and without load with the help of the data received from the Load Sensors (1).
(d) Analysing the risk events occurred while the vehicle was in motion with the data received from the on-board cameras (5) and the accelerometer
(2).
(e) Formulating a datasheet with the above data collected to help a user
ascertain whether a driver would be eligible for credit or loan. The more

the vehicle has travelled with load and low risk events, the better the driver may be associated with credit worthiness.
3. A system for assessing a driver's credit worthiness as claimed in claim 1 wherein the load sensor (1) comprises of strain gauge type or piezo-electric type load cells.
4. A system for assessing a driver's credit worthiness as claimed in claim 1 wherein the accelerometer (2) is a Piezo-electric or piezo-resistive accelerometer.
5. A system for assessing a driver's credit worthiness as claimed in claim 1
wherein the sensor (3) is a GPS sensor.
6. A system for assessing a driver's credit worthiness as claimed in claim 1 wherein the IOT is a GSM chip (4) capable of enabling 2G/3G/4G machine-to-machine (M2M) communication by transferring the on-board sensor data to the server (6).
7. A system for assessing a driver's credit worthiness as claimed in claim 1 wherein the cloud based server system is capable of registering sensor and camera input parameters by integrating a GSM module (4) with the data fetched from the sensors to the server (6).
8. A system for assessing a driver's credit worthiness as claimed in claim 1 wherein the load cell (1) and GPS sensors (3) are sampled every 5 seconds to calculate the earnings only during the presence of a passenger or goods on board.
9. A system for assessing a driver's credit worthiness as claimed in claim 1
wherein the driver earnings are calculated using the following formula:
Driver earnings/day = Rupees/km * distance travelled (in kms) * passenger load (Yes or No/0 or 1)

10.A system for assessing a driver's credit worthiness as claimed in claim 1 wherein the distance travelled between two samples is calculated using the Haversine formula:
Delta lat = lat2 - latl (in radians)
Delta long = long2 - longl(in radians)
A = sin(delta lat/2)A2 + cos (latl in radians) *cos(lat2 in
radians) *sin(delta long/2) A2
C = 2*asin(sqrt(A))
R = radius of earth in kms = 6372.8 kms
Distance travelled = R*C
11.A system for assessing a driver's credit worthiness as claimed in claim 1
wherein the object detection and distance determination is done using
image processing algorithms. 12.A system for assessing a driver's credit worthiness as claimed in claim 1
wherein the system detects the real time driving behaviour and potential
risk of a driver. 13.A system for assessing a driver's credit worthiness as claimed in claim 1
wherein the system collects location, load and driving behaviour data at
fixed intervals.