A vehicle insurance policy may include several types of coverage: bodily injury liability, property damage liability, medical payments, uninsured motorist protection, collision coverage and comprehensive (physical damage). Biographical information is often used as a proxy for actual driving information to determine insurance risk scores for a policy. In this way, in lieu of twenty four hour monitoring of driving behavior, insurance companies have correlated biographical indicators with the chances of a claim (expected losses) being filed.
While these biographical indicators may statistically provide accurate information to an insurance company from a business sense, it may not provide the granularity to accurately assess the risk of a particular driver.
Accordingly, methods and apparatus using telematics are described for telematics based underwriting.
A system is disclosed for determining risk associated with a driver. The disclosed system comprising a computer memory for receiving biographical information associated with one or more drivers, the biographical information including an expected total mileage driven by a vehicle; the memory further configured to store loss data; a processor configured to generate an initial risk assessment based on at least the expected total mileage driven; the processor further configured to generate an insurance quote based at least in part on the initial risk assessment; a receiver, configured to receive from a telematics device, telematics data indicating at least vehicle location and speed and a time stamp; the processor further configured to determine, based at least in part on the telematics data, a plurality of relativity factors; the processor configured to calculate the product of the plurality relativity factors with a starting discount and compare the product with a predetermined threshold; and the processor further configured to adjust pricing information based on the comparison of the product of the relativity factors with the predetermined threshold.
A more detailed understanding may be had from the following description, given by way of example in conjunction with the accompanying drawings wherein:
Disclosed herein are processor-executable methods, computing systems, and related technologies for telematics based underwriting.
In one example, the processor-executable methods and computing systems are configured to use relativity information in the underwriting process. The system may determine expected losses based on loss experience and actual driving behavior.
During a registration phase for vehicle insurance, an account template is opened for a potential customer. The insurance company or an insurance agent may request biographical data, for example via a webpage, to populate information in the account template. The biographical data, may include: name, age, gender, occupation, vehicle, driving history, geographical location, grades (if the driver is a student), and frequency of use of the vehicle. Once the account template is completed, the biographical data stored in the account template is formatted and stored in a database. The system, using software based statistical analysis (e.g. regression analysis) compares the biographical data with actuarial data stored in the system. This actuarial data may include demographic data related to insurance pricing and may include loss data. The system, using the results of the statistical analysis generates an initial risk assessment for the account. As an example, the risk assessment may be categorized by vehicle or by driver. This risk assessment may ultimately be used to determine whether to offer coverage, the rate associated with the coverage, and discount or penalize the rate associated with the coverage.
As will be greater described in detail below, telematics data is collected from the vehicle, providing the insurance company with information such as speed, acceleration, deceleration, left turns, right turns, braking, time of day, mileage, and location.
The telematics data may be analyzed based on stored demographic information, to determine a plurality of relativity factors. These relativity factors may be based on speeding, braking, acceleration, turns, mileage, time of day analysis, driving location risk, distracted driving, hot spot driving, and the types of weather during driving. Further these relativity factors may be numeric value(s) for a type of measured driving behavior. The relativity factor may be relative to other drivers within the same demographic, driving on the same or similar roads under the same or similar conditions, or to the posted speed limit, or driving regulations. Based on the determined relativity factors, the system can determine a discount relativity factor. A computer system then uses a multivariate analysis to generate an adjusted risk score based on the results of this analysis. This risk score may be used to determine adjusted rates. The adjusted rates may be for an overall policy adjustment or for specific coverage, such as for property damage liability, medical payments, uninsured motorist protection, collision coverage, and comprehensive physical damage more accurately.
The telematics data may be received for a predetermined time period. In one example, a telematics device may be installed in a vehicle for a six month period over which data is collected. Because of seasonal changes in driving patterns, (e.g. for students no school during summer time), the DCU 110 may be configured to account for these differences and compensate for seasonal variations by weighting the time frame of the use, using a seasonality factor. Alternatively, the telematics device may be installed for a full year, or be permanently installed. In another embodiment, a software application installed on a mobile phone or other wireless device may be configured to generate the telematics data and communicate with the system 100.
The one or more telematics devices associated with the vehicle 140 may communicate with a satellite, Wi-Fi hotspot and even other vehicles. The telematics devices associated with the vehicle 140 report this information to the DCU 110. As will be described in greater detail hereafter, the DCU 110 may transmit this telematics data to the DPU 170 which may be configured to use telematics data to generate relativity factors.
The web site system 120 provides a web site that may be accessed by a user device 130. The web site system 120 includes a Hypertext Transfer Protocol (HTTP) server module 124 and a database 122. The HTTP server module 124 may implement the HTTP protocol, and may communicate Hypertext Markup Language (HTML) pages and related data from the web site to/from the user device 130 using HTTP. The web site system 120 may be connected to one or more private or public networks (such as the Internet), via which the web site system 120 communicates with devices such as the user device 130. The web site system 120 may generate one or more web pages that may communicate the web pages to the user device 130, and may receive responsive information from the user device 130.
The HTTP server module 124 in the web site system 120 may be, for example, an APACHE HTTP server, a SUN-ONE Web Server, a MICROSOFT Internet Information Services (IIS) server, and/or may be based on any other appropriate HTTP server technology. The web site system 120 may also include one or more additional components or modules (not depicted), such as one or more load balancers, firewall devices, routers, switches, and devices that handle power backup and data redundancy.
The user device 130 may be, for example, a cellular phone, a desktop computer, a laptop computer, a tablet computer, or any other appropriate computing device. The user device 130 includes a web browser module 132, which may communicate data related to the web site to/from the HTTP server module 124 in the web site system 120. The web browser module 132 may include and/or communicate with one or more sub-modules that perform functionality such as rendering HTML (including but not limited to HTML5), rendering raster and/or vector graphics, executing JavaScript, and/or rendering multimedia content. Alternatively or additionally, the web browser module 132 may implement Rich Internet Application (RIA) and/or multimedia technologies such as ADOBE FLASH, MICROSOFT SILVERLIGHT, and/or other technologies. The web browser module 132 may implement RIA and/or multimedia technologies using one or web browser plug-in modules (such as, for example, an ADOBE FLASH or MICROSOFT SILVERLIGHT plug-in), and/or using one or more sub-modules within the web browser module 132 itself. The web browser module 132 may display data on one or more display devices (not depicted) that are included in or connected to the user device 130, such as a liquid crystal display (LCD) display or monitor. The user device 130 may receive input from the user of the user device 130 from input devices (not depicted) that are included in or connected to the user device 130, such as a keyboard, a mouse, or a touch screen, and provide data that indicates the input to the web browser module 132.
The example architecture of system 100 of
Each or any combination of the modules shown in
The web browser window 200 may include a control area 265 that includes a back button 260, forward button 262, address field 264, home button 266, and refresh button 268. The control area 265 may also include one or more additional control elements (not depicted). The user of the user device 130 may select the control elements 260, 262, 264, 266, 268 in the control area 265. The selection may be performed, for example, by the user clicking a mouse or providing input via keyboard, touch screen, and/or other type of input device. When one of the control elements 260, 262, 264, 266, 268 is selected, the web browser module 132 may perform an action that corresponds to the selected element. For example, when the refresh button 268 is selected, the web browser module 132 may refresh the page currently viewed in the web browser window 200.
While the below examples describe a scenario of a new customer registering for insurance and then having the pricing information adjusted based on telematics data, the systems and methods described herein may be applied to current and former customers that are looking to renew their coverage. In this scenario, the biographical information may already be stored on the insurance server 180, and the DPU 170 may access this information directly.
The registration phase is used to generate an initial risk assessment. During the registration phase, the system 100 receives biographical information about each of the drivers that may be associated with the user's account as well as information about the vehicles for which coverage is requested. With millions of accidents each year, a large amount of data is available on factors that may affect the likelihood of an accident as well as the severity of the accident. The database 176 associated with the DPU 170 contains information regarding accident information. The DPU 170, using a multivariate analysis, generates the initial driver assessment based on the provided biographic information verses the factors stored in the database 176.
The DPU 170 may perform a correlative analysis on the entered biographical information to develop the initial risk assessment which may be based in part on the expected speeding, the expected acceleration, the expected turns, the expected braking, the expected mileage driven, the times of day driven, etc. The list above is by no means exhaustive. Based on the entered biographical information, the DPU 170 may also be configured to generate an expectation on time spent in low risk, medium risk, and high risk locations (other than the specific expected locations.) The RPU 160 may use this information to generate pricing information. For example, the RPU 160 may adjust the rate associated with an account, it may credit or debit a rate and/or to determine adjusted pricing information.
The inside of vehicle 140 may comprise a plurality of electronics devices that may communicate information to the telematics device. Most vehicles include at least one microprocessor and memory that connects to each individual electronic device. For example, there may be electronic devices associated with the seats, A/C units, global positioning satellite (GPS)/stereo system, DVD unit, and BLUETOOTH equipment. The microprocessor may also be in communication with the headlights, engine, traffic signals, rear view mirror, rearview cameras, cruise control, braking system and inner workings of the vehicle 140. There may also be additional devices such as multiple mobile phones brought by passengers into a vehicle. The telematics device is configured to receive information from the electronics in the vehicle 140. For example, the telematics device is configured to receive data concerning, speed, acceleration, turns, braking, location, seat settings, lane changes, radio volume, window controls, vehicle servicing, number of cellular devices in a vehicle, proximity to other vehicle's, etc. The telematics device may be configured to transmit this information directly to the DCU 110.
The DCU 110 may format this information and transmit it to the DPU 170. Once the account has been activated, the DPU 170 may be configured to use this information to determine the relativity factors associated with each vehicle.
The telematics device may be configured to record telematics data periodically as well as based on a trigger. Based on this information, the DPU 170 may be configured to determine a plurality of relativity factors for the measured data categories. In one embodiment, the relativity factors may be based on predetermined road segments.
For example, the DPU 170 may also be configured to categorize portions of road as road segments, wherein road segments may be predetermined lengths of road. As a preliminary basis, the DPU 170 may label a first category of roads “highways,” including: interstates, U.S. highways, limited-access highways as “highways” or “primary roads”. The DPU 170 may label a second category of roads as “urban,” including: secondary roads, and local roads of high importance. The DPU 170 may label a third category of roads as “other,” including: local roads of minor importance, alleys, other unpaved roads or footpath.
Alternatively or additionally, the DPU 170 may be configured to determine the relativity factors in relation to nearby drivers or drivers on similar roads under similar conditions.
In a first example, the DPU 170 may be configured to determine a driving location relativity factor. For example, the driving location relativity factor may credit or penalize a driver for driving in locations more or less risky than their home address. The database 176 of the DPU 170 may generate a driving location risk index (DLRI), wherein the DLRI comprises rankings of each driving location, a vehicle may encounter. The DLRI may be based on a predetermined area. This granularity may be adjusted based on the available telematics and loss data. As one example, where allowable by law, the DLRI may be categorized by zip code. After receiving telematics data from the telematics device of vehicle 140, the DPU 170 may be configured to compare the driving location, with the DLRI to determine the relative risk of the locations.
For example, the DPU 170 may calculate the relative risk of the reported locations actually driven compared to the expected home location according to the procedure described below. The DPU 170 may determine the total number of miles driven by zip code. Next, the DPU 170 may calculate a state adjustment factor. The state adjustment factor may be calculated, e.g. according to the equation 1:
State adjustment factor=State Avg. Premium/State Avg. Base Rate. (EQ. 1)
Wherein the state adjustment factor is based on bodily injury, property damage, comprehensive and collision coverage factors. The DPU 170 may use the state adjustment factor may be used to calculate adjusted base rates by zip code, based on Eq. 2 below:
Adjusted Base Rates by Zip Code=State Adjustment Factor×Base Rate (EQ. 2)
The DPU 170 may use this information to generate adjusted base rates for each of the locations. An example of weighted average rates, based on the driving location, is shown in Table 1, below.
Based on the percentage of miles driven in each zip code, a rate is determined. The driving location relativity is determined according to the Eq. 3.
Driving location relativity=Sqrt(wtd avg of rates/rate of home zip (EQ. 3)
Wherein a DLRI>1 indicates that the vehicle is driven in riskier areas than the home location. And a DLRI<1 indicates that the vehicle is driven in less risky areas than the home location.
The DPU 170 may further be configured to generate a braking relativity factor. To generate a braking relativity factor, the DPU 170 must determine if a predetermined condition is satisfied such that a braking event is declared. For example, the DPU 170 may declare a braking event based on a rate deceleration or the amount of pressure applied to a brake. The database 176 of the DPU 170 may further be configured to store braking benchmarks for each type of road segment. An example of the braking benchmarks is shown below in Table 2.
Based on received telematics data, the DPU 170 determines the frequency and location of each braking event. This information is compiled in the database 176, and the DPU 170, then determines the amount of braking events per mile for each type of road segment and the overall proportion of braking for each road segment. Table 3 shows an example of compiled braking data.
For each type of road segment, an index is determined, wherein the index=measured/benchmark. For the example above, HW_Index=0.12/0.01, UR_Index=0.29/0.07=4.1, and OT_Index=0.32/0.03.
The DPU 170 may be configured to calculate an overall breaking index by averaging each of the braking indices weighted by the proportion of miles driven on each road. In the example above, the overall braking index may be calculated as follows:
Overall Braking Index=HW_Index*prop_miles_driven—HW+UR_index*prop_miles_driven—UR+OT_Index*prop_miles_driven_Other. (EQ. 4)
The DPU 170 may be configured to rescale the overall braking index and center it around 1. This overall braking index may be scaled according to the following equation:
Scaled Braking Index=(Overall_Braking Index−mean of the distribution)/(standard deviation of the distribution)+1 (EQ. 5)
Wherein the mean and standard deviation of the distribution come from a lookup table
The system 100 may be able to adjust pricing data with or without loss data. For example, in absence of enough credible loss data from telematics devices, (enough losses in the data to have desired statistical power), the system 100 may determine an expected loss value, also known as Expected Pure Premium (EPP) to calculate a braking relativity factor, wherein the EPP is calculated based on conventional class plan variables. The EPP may then be regressed on the telematics variables like braking, speeding etc. in a multivariate scenario to derive coefficients for these telematics variables. In another embodiment, the system 100 may use a univariate analysis and the EPP may be used to calculate the slope for the telematics variable. Using a look up table, stored in database 176, the DPU 170 may map the scaled braking index to a braking relativity factor. An example of mapping a scaled braking index to a braking relativity factor is shown in Table 4 below. According to the Table 4, an expected pure premium may be used.
The DPU 170 may further be configured to determine a speeding relativity factor. The database 176 of the DPU 170 may be preconfigured to store a speed benchmark for each road segment. Table 5, below shows an example of a speed benchmark, using the same segments determined for the braking benchmark. This is used as an illustrative example only. In another embodiment, the road segments for speed may be determined based on posted speed limits, or measured clustered driving patterns.
After receiving the telematics data, the DPU 170 may be configured to calculate the proportion of miles driven 20 mph over the speed benchmark, 10 to 20 mph over the speed benchmark, 1 to 10 mph over the speed benchmark and 0 mph over the speed benchmark for each of the types of road segment. Further, the DPU 170 may be configured to assign weights based on the variance from the speed benchmark. An example for highway segments is shown in Table 6, below. While the table below only shows weights for speed above the speed benchmark, it may also include weights for speeds below the speed benchmark.
The DPU 170 calculates a speeding index for each road segment by multiplying the risk weight of each speed grouping (e.g. HW—20 mphover) by the proportion of miles within that bucket. For example, based on the three equations given below:
HW_Index=Highway—20_mph_over_prop*wt+Highway—10to20_mph_over_prop*wt+Highway—0to10_mph_over_prop*wt+Highway—0_over*wt (EQ. 6)
UR_Index=UR—20_mph_over_prop*wt+UR—10to20_mph_over_prop*wt+UR—0to10_mph_over_prop*wt+UR—0_over*wt (EQ. 7)
OT_Index=OT—20_mph_over_prop*wt+—OT—10to20_mph_over_prop*wt+OT—0to10_mph_over_prop*wt+OT—0_over*wt (EQ. 8)
The DPU 170 may further generate an average of the speeding indices weighted by proportion of miles driven on each road segment to determine an overall speeding index, wherein:
Overall_Speeding_Index=HW_Index*prop_miles_driven—HW+UR_Index*prop_miles_driven_Urban+OT_Index*prop_miles_driven_Other (EQ. 9)
The DPU 170 may further be configured to determine an overall speeding index that is used to determine the speeding relativity factor. Table 7 shows an overall speeding index mapped to a speeding relativity factor.
The DPU 170 may further be configured to determine a mileage relativity factor. The mileage relativity factor may be based on an expected mileage value entered by the user during the registration phase. The expected mileage is compared with the measured mileage. The DPU 170 may mitigate the effect of the relativity factor, for example by operating on the result with a function. As an example, the mileage relativity may be calculated as follows, using a square root function to mitigate the effect:
Mileage relativity=SQRT(mileage factor based on actual miles driven/mileage factor based on reported miles) (EQ. 10)
The DPU 170 may further be configured to determine a time of day relativity factor. Based on loss data, the DPU 170 may categorize time segments as high risk, low risk and moderate risk. The DPU 170 may measure the relative risk of driving at certain times of day. The DPU 170 may weight each of the times of day, wherein the weighting rewards low risk miles while incrementally penalizing moderate and high risk miles. Based on the received telematics data, the DPU 170 may further calculate the proportion of miles driven within each time of day segment. Table 8, below, shows an example of time of day weighting.
The DPU 170 may then calculate a time of day (TOD) risk index based on the mileage weighted average of TOD risk. The TOD risk index is mapped to a TOD relativity factor, using a lookup table. Table 9 shows a (TOD) risk index and TOD relativity factor based on the example above.
The DPU 170 may transmit the relativity factors to the RPU 160. The RPU 160 may be configured to adjust the rate, or provide a discount or surcharge based on the relativity factors according, for example, to the equation below:
Discount relativity=starting discount*driving location relativity*braking relativity*speeding relativity*mileage relativity*time of day relativity (EQ. 11)
The system 100 may further be configured to determine whether the vehicle 140 is a self-driving vehicle, in which an on-board computer operates the vehicle 140. In this case, the effect of the driving time of day or any other factor may be mitigated when determining the pricing information.
The system 100 uses the biographical information provided in web pages 302-1302 as a baseline for generating the initial pricing information. However, using the methods described above and the received telematics data, provided by the telematics device, the system 100 may refine the pricing information by adjusting the rate, providing a credit or surcharge, or rejecting a renewal. In one embodiment, the RPU 160 may access the information stored in the DPU 170 and the determined discount relativity, and use a software based algorithm to determine a discount.
For example, the starting discount may be 10%, and if the product of the direct and indirect exposure ratings with the weighting factors >1, the system 100 may determine the driver is not eligible for a discount.
In one scenario, the system 100 may only receive telematics data for a fixed time period. In this scenario, the RPU 160 may be configured to compensate for the limited duration of the telematics data using a seasonality factor. For example, if the telematics data is received from September-December, and the biographical information indicates one of the insured drivers attends college away from home, RPU 160 may be configured to use the seasonality factor to adjust the pricing information to account for the lack of information transmitted regarding that driver. Conversely, under the same scenario, if the readings were taken during the summer, when the student was home, the telematics data may be skewed the other way. Accordingly, the RPU 160 may use the seasonality factor to account for that.
The memory device 1520 may be or include a device such as a Dynamic Random Access Memory (D-RAM), Static RAM (S-RAM), or other RAM or a flash memory. The storage device 1516 may be or include a hard disk, a magneto-optical medium, an optical medium such as a CD-ROM, a digital versatile disk (DVD), or Blu-Ray disc (BD), or other type of device for electronic data storage.
The communication interface 1522 may be, for example, a communications port, a wired transceiver, a wireless transceiver, and/or a network card. The communication interface 1522 may be capable of communicating using technologies such as Ethernet, fiber optics, microwave, xDSL (Digital Subscriber Line), Wireless Local Area Network (WLAN) technology, wireless cellular technology, BLUETOOTH technology and/or any other appropriate technology.
The peripheral device interface 1512 may be an interface configured to communicate with one or more peripheral devices. As an example, the peripheral device may communicate with an onboard diagnostics (OBD) unit that is associated with a vehicle. The peripheral device interface 1512 may operate using a technology such as Universal Serial Bus (USB), PS/2, BLUETOOTH, infrared, serial port, parallel port, and/or other appropriate technology. The peripheral device interface 1512 may, for example, receive input data from an input device such as a keyboard, a mouse, a trackball, a touch screen, a touch pad, a stylus pad, and/or other device. Alternatively or additionally, the peripheral device interface 1512 may communicate output data to a printer that is attached to the computing device 1510 via the peripheral device interface 1512.
The display device interface 1514 may be an interface configured to communicate data to display device 1524. The display device 1524 may be, for example, an in-dash display, a monitor or television display, a plasma display, a liquid crystal display (LCD), and/or a display based on a technology such as front or rear projection, light emitting diodes (LEDs), organic light-emitting diodes (OLEDs), or Digital Light Processing (DLP). The display device interface 1514 may operate using technology such as Video Graphics Array (VGA), Super VGA (S-VGA), Digital Visual Interface (DVI), High-Definition Multimedia Interface (HDMI), or other appropriate technology. The display device interface 1514 may communicate display data from the processor 1518 to the display device 1524 for display by the display device 1524. As shown in
An instance of the computing device 1510 of
Although
As described above, the relativity factors may be based on different units of area. In another example, the relativity factors may be determined relative to road segments travelled (e.g. braking per road segment).
As shown in
Regarding
The system 100 may further include a user transmission device (not pictured) wherein the user transmission device may communicate insurance information, including pricing information, contractual information, information related to the telematics program, and other notifications. A user transmission device may include one or more modes of communication to reach a potential customer, current customer, or past customer or other similar user. For example, the user transmission device may be coupled with a printing device that is automatically mailed to the user. In another embodiment, the user transmission device may be coupled to a device to generate automatic telephone calls, or “robo-calls,” or other similar communication mediums to communicate with the user. The user transmission device may further be configured to send e-mails to a user. The user device may further be configured to communicate via social media.
The system 100 may communicate this information during a renewal period. Additionally, the system 100 may be configured to proactively communicate this information and/or adjust the pricing information based on exposure changes determined by the system 100 that may occur within or outside of the renewal period.
The multivariate predictive model(s) may include one or more of neural networks, Bayesian networks (such as Hidden Markov models), expert systems, decision trees, collections of decision trees, support vector machines, or other systems known in the art for addressing problems with large numbers of variables. In embodiments, the predictive models are trained on prior data and outcomes using a historical database of insurance related data and resulting correlations relating to a same user, different users, or a combination of a same and different users. In embodiments of the present invention, the predictive model may be implemented as part of the DPU 170 or RPU 160 described with respect to
As used herein, the term “processor” broadly refers to and is not limited to a single- or multi-core processor, a special purpose processor, a conventional processor, a Graphics Processing Unit (GPU), a digital signal processor (DSP), a plurality of microprocessors, one or more microprocessors in association with a DSP core, a controller, a microcontroller, one or more Application Specific Integrated Circuits (ASICs), one or more Field Programmable Gate Array (FPGA) circuits, any other type of integrated circuit (IC), a system-on-a-chip (SOC), and/or a state machine.
As used to herein, the term “computer-readable medium” broadly refers to and is not limited to a register, a cache memory, a ROM, a semiconductor memory device (such as a D-RAM, S-RAM, or other RAM), a magnetic medium such as a flash memory, a hard disk, a magneto-optical medium, an optical medium such as a CD-ROM, a DVD, or BLU-RAY disc, or other type of device for electronic data storage.
Although the methods and features described above with reference to
The following documents are incorporated herein by reference as if fully set forth: U.S. application Ser. No. 14/145,142, titled SYSTEM AND METHOD FOR DETERMINING DRIVER SIGNATURES filed Dec. 31, 2013; U.S. application Ser. No. 14/145,165, titled SYSTEM AND METHOD FOR EXPECTATION BASED PROCESSING filed Dec. 31, 2013; and U.S. application Ser. No. 14/145,181, titled SYSTEM AND METHOD FOR DESTINATION BASED UNDERWRITING filed Dec. 31, 2013. Each of the applications shares common inventorship with the present application and are being filed concurrently.