Patent Application
20140136242
The present application relates generally to insurance and, more specifically, to systems and methods for collecting and processing data for insurance rating purposes.
In insurance industries, such as property/casualty, liability, life, and health insurance industries, insurance providers generally seek to determine an insurance policy premium that is appropriate given the risk of losses (e.g., property theft, property damage, health issues requiring medical treatment, death, etc.) for the individual owning the policy. For purposes of making this determination, it is well understood that behaviors of an individual can exert a great influence on the probability that the individual experiences a loss that is recognizable under the policy. For example, an individual who smokes cigarettes on a regular basis is much more likely to incur large medical bills as compared to a non-smoker. In some circumstances, behaviors such as this must be learned based on information provided by the insurance policy holder, or future insurance policy holder, in response to questions from the insurer (e.g., questions regarding how much the individual smokes cigarettes, drinks alcoholic beverages, etc.). Generally, individuals demonstrating behaviors corresponding to a lower risk of loss (“low-loss behaviors”) may be assigned a more positive rating, and may therefore be offered lower premiums for a given level of coverage. Conversely, individuals demonstrating behaviors corresponding to a higher risk of loss (“high-loss behaviors”) may be assigned a more negative rating, and may therefore be offered higher premiums for the same level of coverage.
Unfortunately, insurers generally have access to a very limited amount of information with respect to policy holder behaviors. Questionnaires provided by insurers to prospective policy holders are typically very limited in scope, as insurers may only be aware of a small subset of the universe of behaviors affecting the risk of loss. Moreover, responses to insurer questionnaires may in some cases be inaccurate. For example, a prospective policy holder may not know the precise number of alcoholic beverages, on average, that he or she imbibes per week. Further, companies providing certain types of insurance policies are at a particular disadvantage in this respect, as behaviors that substantially affect the likelihood of some types of losses are generally not well characterized, and may be difficult to assess based on questionnaires. For example, companies providing homeowners insurance may be unaware of many behaviors affecting the risk of losses within the home, and/or may find it difficult to determine whether a policy holder exhibits those behaviors. As a result, insurance ratings and premiums determined for the policy holder may be poorly correlated to the policy holder's risk of loss.
The disclosed system collects data from one or more devices of a residence associated with a current or potential insurance policy holder, and analyzes the collected data (generally referred to herein as “in-home data”) in order to rate the policy holder for insurance purposes, such as determining the policy holder's insurance premium. As used herein, “residence” does not necessarily refer to a legal residence, and may include a real property that an individual owns, rents, leases, etc., but is not necessarily inhabited. The in-home data may be based on information generated by any of various devices found in home-centered systems, such as home automation and monitoring systems, and may be indicative of behaviors and/or conditions within the residence. For example, motion sensor, window sensor, and/or door sensor devices may generate data indicating times during which a residence is occupied, and home security system devices may indicate an occupant's habits with respect to arming the system. Once collected, the in-home data is analyzed to determine whether the in-home data is indicative of behaviors and/or conditions that are known to raise or lower the risk of a recognizable loss under the individual's insurance policy. For example, if it is known that certain usage patterns relating to home security systems (e.g., arming a system with at least a threshold frequency, arming a system at particular times of day, etc.) statistically correlate to a lower risk of in-home losses, then data generated by a home security device in the policy holder's home may be analyzed to determine whether any of those usage patterns exist. If the in-home data is determined to correspond to a behavior or condition known to lower the risk of loss, the policy holder's insurance premium may be lowered accordingly, for example.
In the embodiment shown in
The processor 52 also controls the actuators 58. The actuators 58 include devices (e.g., devices including switches/relays, motors, valves, etc.) for controlling various devices within the residence 20, such as light fixtures, fans, televisions, appliances, outlets, door locks, water shutoff valves, and/or automated blinds, for example. A memory (not shown in
In the embodiment shown in
In other embodiments, different devices or systems are located at the residence 20, and/or the devices at the residence 20 are coupled to the network 40 in different ways than shown in
Further, in some embodiments, the gateway 60 is coupled to the network 40 via one or more in-home networks (not shown in
Still further, in some embodiments, some or all of the in-home data associated with the residence 20 is sent to the insurer's computer system 30 via a third party, rather than directly from the residence 20. For example, a server of a home security system provider (not shown in
Referring again now to the embodiment shown in
The insurance rating server 70 stores the received in-home data in a memory 72, where the in-home data can be retrieved at a later time for processing. The insurance rating server 70 is also configured to retrieve correlation data stored in a memory 74. In other embodiments, the memory 72 and/or the memory 74 is/are instead located outside of the insurer's computer system 30, and is/are accessible by the insurance rating server 70 via a network such as the network 40. The correlation data stored in the memory 74 may include data modeling correlations between (a) usage patterns of in-home devices, and/or patterns relating to conditions monitored by in-home devices, and (b) likelihoods of incurring recognizable losses under the policy holder's policy. The insurance rating server 70 may be configured to analyze the in-home data stored in memory 72 using one or more of these correlation models in order to determine a risk rating, or a parameter corresponding to a risk rating (e.g., a change in an insurance premium). As an example in which a relatively simple correlation model is used, the insurance rating server 70 may compare the number of hours that the alarm 54 has been armed in a given month (as determined based on the in-home data received from processor 52) with one or more ranges of hours identified by the correlation data (e.g., 0-100 hours, 101-200 hours, etc.), and determine a risk indicator that the correlation data indicates as being associated with the range that matches the in-home data. To this end, the memory 74 may include a relational database, with each hour range corresponding to an indicator of a loss likelihood, for example. The insurance rating server 70 may then determine an insurance premium adjustment that corresponds to the identified risk indicator, such as a discount (5%, 10%, etc.) if a “low” risk has been identified. Alternatively, the insurance rating server 70 may determine a different benefit, such as an offer of a home security or home automation product, in order to reward the low-risk behavior of the policy holder without necessarily adjusting the policy holder's premium.
Alternatively, or additionally, the correlation data stored in the memory 74 may include more complex models or algorithms that depend on multiple types of data, generated by multiple in-home devices, in order to relate in-home data to risk of loss. Various example correlations are described below in connection with
The correlation data stored in the memory 74 may be based on manually entered information, or may be “learned” by the insurance rating server 70 (or another server not shown in
In some embodiments, the insurance rating server 70 checks whether the policy holder residing at the residence 20 has opted in to a discount/rewards program before utilizing his or her in-home data to determine an insurance rating. Alternatively, the devices within the residence 20 may not even be configured to communicate with the insurer's computer system 30 unless the policy holder has already opted in to the program, in which case the insurance rating server 70 may not need to determine whether the individual has opted in before utilizing the in-home data. In still other embodiments, the insurance rating system 100 utilizes in-home data of the policy holder without requiring the policy holder to opt in or agree to a specific program.
As noted above, information from various other types of in-home devices and systems may be utilized for insurance rating purposes.
Each of the in-home data sources 120A-120E in
The insurance rating server 130 may operate directly on the data provided by data sources 120A-120E, or may operate on other data that is generated based on the data from data sources 120A-120E. For example, the insurance rating server 130 may process the data from in-home data sources 120A-120E and convert it to a particular format (e.g., for efficient storage), and later utilize the modified data for insurance rating purposes.
In addition to receiving data from the in-home data sources 120A-120E over the network 140, the insurance rating server 130 receives data from an additional data source 142 coupled to the insurance rating server 130. The data source 142 provides information about external factors regarding the residence associated with in-home data sources 120A-120E (e.g., a home address, a crime rate associated with a geographic area that includes the residence, or other environmental factors), and/or the policy holder associated with the residence (age, gender, etc.), that influence or may influence the risk of loss under a policy. In other embodiments, the data source 142 is instead coupled to the network 140, and the insurance rating server 130 receives data from the data source 142 via the network 140. In still other embodiments, the insurance rating system 100 does not include the data source 142.
The insurance rating server 130 also receives data from a correlation data source 150 coupled to the insurance rating server 130. The insurance rating server 130 and correlation data source 150 may be similar to the insurance rating server 70 and memory 74 of
Based on the determined insurance rating, the insurance rating server 130 generates an indication 160 of a premium adjustment (e.g., a premium discount in response to determining that the policy holder has a reduced risk of loss). In other embodiments, the insurance rating server 130 generates an indication of an incentive or reward other than a premium adjustment, such as a home security or automation device, a carbon monoxide detector, a smart appliance, etc. The indication may be information displayed to an operator of the insurer's computer system, data provided to a software module within the insurer's computer system, or a printable statement file which can be delivered to the policy holder, for example. The adjustment to the premium, and/or the total premium including the adjustment, may then be communicated to the policy holder.
The insurance rating system 100 of
In one example embodiment, the source 120B or the source 120D includes a tilt sensor detecting whether a garage door of the residence is open or closed, and the correlation data stored in the correlation data source 150 represents a correlation model under which a garage door that is more frequently left open (or left open at certain times of day or night, etc.) corresponds to a higher risk of loss. The insurance rating server 130 may therefore use data from the tilt sensor to determine a risk of loss and/or insurance premium adjustment according to the correlation model.
In another example embodiment, the source 120B includes one or more motion sensors, door sensors, and/or window sensors, and the correlation data stored in the correlation data source 150 represents a correlation model under which the percentage of time that a policy holder is home is inversely proportional to the policy holder's risk of a loss in the home (e.g., due to the individual being able to address any issues within the home, such as fire or water damage, as opposed to someone who is rarely home). For example, a vacation home or a vacant home may be at higher risk than a primary residence. The insurance rating server 130 may therefore use data from the motion sensors, door sensors, and/or window sensors to determine a risk of loss and/or insurance premium adjustment according to the correlation model.
In yet another example embodiment, the source 120C includes gas, water, and/or electricity meters that detect and indicate usage of the respective utilities, and the correlation data stored in the correlation data source 150 represents a correlation model under which a utility usage above a certain threshold (e.g., a determined average utility usage in the neighborhood of the policy holder's residence) corresponds to a higher risk of loss in the home. Alternatively (or additionally), the correlation data may represent a correlation model under which a very low electricity usage, which may indicate that the residence is not occupied, corresponds a higher risk of loss. The insurance rating server 130 may therefore use data from the utility meter(s) to determine a risk of loss and/or insurance premium adjustment according to the correlation model.
In still another example embodiment, devices of a home monitoring system (e.g., including the sources 120A and 120B) may generate information indicating the state or “health” of the system, such as whether the system is sufficiently powered (e.g., by batteries in the devices), kept online and connected, etc., and the correlation data stored in the correlation data source 150 represents a correlation model under which an “unhealthy” home monitoring system (e.g., one in which low batteries are not quickly replaced, etc.) corresponds to a higher risk of loss in the home. For example, a policy holder who is slow to replace batteries may also be less likely to practice other behaviors that tend to prevent losses, such as cleaning gutters, cleaning the lint out of a dryer vent, etc. The insurance rating server 130 may therefore use data from devices of the home monitoring system to determine a risk of loss and/or insurance premium adjustment according to the correlation model.
It is understood that the above examples are not exclusive, and that more than one such embodiment may coexist within a single insurance rating system.
The method 200 receives in-home data that is generated by, or based on information generated by, a device located at a residence of an individual (block 210). The residence may be a single-family home, townhome, condominium, or apartment, for example, and the individual may be a current policy holder or a potential policy holder (e.g., an individual to whom an insurance quote may be offered). The device generates in-home data indicative of a utilization of the device (e.g., data indicative of whether a security device is armed and the alarm mode, data indicative of usage of a light fixture or a smart appliance, etc.), and/or indicative of a condition monitored by the device (e.g., data indicative of sensed motion, data indicative of door or window positions, data indicative of the presence of smoke or carbon monoxide, data indicative of an amount of water, gas or electricity being used, etc.). For example, the in-home data may indicate one or more time periods during which the device was utilized, and/or during which the device detected a monitored condition. The device may be a device in the home automation and monitoring system 50 of
The in-home data is received via a communication network, such as the network 40 of
The method 200 also determines an insurance premium adjustment for the individual, based on the in-home data received at block 210 (block 220). The premium may be a monthly, quarterly, or annual premium, for example, and the premium may be for property/casualty insurance, homeowners insurance, or a different type of insurance. In some embodiments, the adjustment can be either a premium discount or “no change,” depending on the in-home data. In other embodiments, the adjustment can be either a discount or a premium penalty/increase. The premium adjustment may be determined for the individual's existing policy (if the individual is a current policy holder), or to be included in a quote (if the individual is a potential policy holder), for example.
In an embodiment, the method 200 determines the insurance premium adjustment based on the received in-home data at least in part by monitoring the in-home data to determine a behavior of one or more occupants of the residence, and determining the insurance premium adjustment based on the determined behavior. As a more specific example, in an embodiment in which the device is a motion sensor, a door sensor, or a window sensor, determining the insurance premium adjustment based on the received in-home data includes determining a pattern of occupancy associated with the residence of the individual (e.g., a percentage of time someone is at the residence, times of day when someone is detected at the residence, etc.), and determining the insurance premium adjustment based on the determined pattern of occupancy.
In an embodiment, the method 200 determines the insurance premium adjustment for the individual at least in part by determining an indication of loss likelihood based on (a) the received in-home data and (b) a known correlation between in-home data associated with one or more insurance policy holders and claims data associated with those insurance policy holders. One such embodiment is described in more detail below in connection with
The method 200 also provides an indication of the insurance premium adjustment determined at block 220 (block 230). The method 200 may provide the indication by displaying information to an operator of the insurer's computer system (e.g., insurer's computer system 30 of
Blocks 210, 220 and 230 may be repeated multiple times. For example, in-home data may be received on a substantially continuous basis (e.g., at various times throughout each day, each week, etc.), and a new insurance premium adjustment may be determined on a periodic basis, or according to any other suitable schedule (e.g., whenever the received in-home data indicates that one or more particular, pre-established conditions have been satisfied). An indication of the new premium adjustment may then be provided each time a new premium adjustment is determined, for example.
In alternative embodiments, the method 200 may include additional blocks not shown in
In yet another example, the method 200 may additionally determine whether the individual has opted into an incentives program. In this embodiment, the method 200 may receive the in-home data (block 210), and/or may determine the insurance premium adjustment based on that data (block 220), in response to determining that the individual has opted into the incentives program. In other embodiments, the method 200 does not determine whether the individual has opted into an incentives program.
In another alternative embodiment, the method 200 determines (block 220) and provides an indication of (block 230) an incentive, other than a premium adjustment, for practicing low-loss behaviors. For example, the method 200 may determine whether to reward the individual with a home automation or security product, and/or which type of product to offer, and provide an indication of any such product.
In the example insurance rating system 300, each of a plurality of residences 310A-310C sends in-home data to an insurer's computer system 320 via a network 330. Each of the residences 310A-310C (e.g., single-family homes, townhomes, condominiums, apartments, etc.) may be the residence of a different insurance policy holder, or, in some embodiments or scenarios, a single insurance policy holder may be associated with two or more of the residences 310. While three residences 310 are shown in
Located at each of the residences 310A-310C are one or more devices capable of generating information relating to usage of those devices, and/or relating to conditions monitored by those devices. For example, each of the residences 310A-310C may include devices similar to the devices of the home automation and monitoring system 50 in
The insurer's computer system 320 includes an insurance rating server 340 configured to receive in-home data from these devices via the network 330. To couple the various data-producing devices in the residences 310A-310C to the network 330, each of the residences 310A-310C may include one or more gateways similar to the gateway 60 of
The insurance rating server 340 collects the in-home data from the residences 310A-310C over time and stores the collected data in a memory 342. The insurance rating server 340 also has access to claims data of the policy holders associated with the residences 310A-310C, which is stored in a memory 344. The claims data indicates actual instances of past losses under the various insurance policies of the policy holders, along with associated information such as the date of the loss, the type of loss, any monies paid to the policy holders due to the loss, etc. The insurance rating server 340 may retrieve the claims data from the memory 344 and the collected in-home data from the memory 342, and utilize the retrieved data to identify particular patterns in the in-home data that exhibit strong correlations with contemporaneous losses identified in the claims data. For example, the insurance rating server 340 may determine that losses occur more frequently in residences that are occupied less often, with occupancy (e.g., as a percentage of time) being determined based on data from motion and/or door sensors. In other embodiments or scenarios, any of the correlations discussed above in connection with the examples provided for
The insurance rating server 340 may generate correlation models based on the identified patterns and correlations, which may then be used with the in-home data of an individual policy holder in order to determine an insurance rating for the policy holder (e.g., for adjusting a premium), as discussed above in connection with
In an embodiment, the insurance rating system 300 is the same as the insurance rating system 10 of
The method 400 receives first in-home data that is generated by, or based on information generated by, a plurality of devices located at a plurality of residences associated with a plurality of individuals (block 410). The first in-home data may include data indicative of utilizations of some or all of the plurality of devices, and/or data indicative of conditions monitored by some or all of the plurality of devices. The in-home data received from each residence may be similar to the in-home data received at block 210 of method 200 in
The first in-home data is received via one or more communication networks, such as the network 330 of
The method 400 also receives claims data associated with the plurality of individuals (block 420). The claims data may be retrieved from a memory such as the memory 344 of
The method 400 also determines, based on the first in-home data received at block 410 and the claims data received at block 420, one or more in-home data patterns corresponding to an increased or decreased probability of a loss recognizable under an insurance policy (block 430). The in-home data patterns may represent patterns of device usage, and/or patterns of sensed/monitored conditions, that can be matched to loss probabilities, and incorporated in a correlation model. For example, the method 400 may determine ranges of device output values that correspond to loss probabilities, scheduling/timing patterns that correspond to loss probabilities, etc.
The method 400 also receives second in-home data generated by, or based on information generated by, a device located at the residence of a current or potential insurance policy holder (block 440). While the first in-home data received at block 410 is used to determine in-home data patterns corresponding to various probabilities of losses in the home (i.e., correlation data), the second in-home data is used to assess the risk of loss associated with a particular (current or potential) policy holder. Block 440 may be similar to block 210 of the method 200 in
The method 400 also determines whether the second in-home data received at block 440 matches at least one of the in-home data patterns determined at block 430. To this end, the second in-home data may be processed utilizing a correlation model generated based on the determined in-home data patterns. For example, in an embodiment in which the method 400 determines (at block 430) that one or more ranges of device output values correspond to an increased or decreased probability of incurring a recognizable loss, the method 400 may determine whether the second in-home data falls within at least one of those ranges of device output values. In other embodiments, the in-home data patterns are more complex, and determining whether the second in-home data matches any of the patterns involves more than simply determining a range within which a device output value falls. For example, the in-home data pattern(s) may reflect multiple parameters (e.g., state of a device or sensed condition, time of the state or condition, etc.), and/or the in-home data pattern(s) may be matched only by satisfying a non-trivial algorithm (e.g., only if a certain state or condition exists at certain times of day, or with a certain frequency, etc.).
The method 400 also, in response to determining at block 440 that the second in-home data matches at least one of the in-home data patterns, determines an insurance premium adjustment for the policy holder (block 460) (e.g., a premium adjustment that corresponds to the matched in-home data pattern(s)). The premium may be a monthly, quarterly, or annual premium, for example, and the premium may be for property/casualty insurance, homeowners insurance, or a different type of insurance. In some embodiments, the adjustment can be either a premium discount or “no change,” depending on the in-home data. In other embodiments, the adjustment can be either a discount or a premium penalty/increase. The premium adjustment may be determined for an existing policy (if for a current policy holder), or to include in a quote (if for a potential policy holder), for example.
The method 400 also provides an indication of the insurance premium adjustment (block 470). Block 470 may be similar to block 230 of
Blocks 410, 420 and 430, and/or blocks 440, 450, 460 and 470, may be repeated multiple times. For example, additional in-home data and claims data associated with the plurality of individuals may be received on a substantially continuous basis (e.g., at various times throughout each day, each week, etc.), and new in-home data patterns may be determined based on the additional in-home data on a continuous basis, a periodic basis, or according to any other suitable schedule. As another example, additional in-home data associated with the policy holder may be received on a substantially continuous basis, with new premium adjustments being determined for the policy holder based on comparisons between the additional in-home data of the policy holder and the original or updated in-home data patterns.
In alternative embodiments, the method 400 may include additional blocks not shown in
In yet another example, the method 400 additionally determines whether the insurance policy holder has opted into an incentives program, in a manner similar to that described above in connection with
In another alternative embodiment, the method 400 determines (block 460) and provides an indication of (block 470) an incentive, other than a premium adjustment, for practicing low-loss behaviors. For example, the method 400 may determine whether to reward the policy holder with a home automation or security product, and/or which type of product to offer, and provide an indication of any such product.
Computer 510 typically includes a variety of computer readable media. Computer readable media can be any available media that can be accessed by computer 510 and includes both volatile and nonvolatile media, and both removable and non-removable media. By way of example, and not limitation, computer readable media may comprise computer storage media and communication media. Computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, random access memory (RAM), read only memory (ROM), EEPROM, FLASH memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by computer 510. Communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media. The term “modulated data signal” means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, radio frequency (RF), infrared and other wireless media. Combinations of any of the above are also included within the scope of computer readable media.
The system memory 530 includes computer storage media in the form of volatile and/or nonvolatile memory such as ROM 531 and RAM 532. A basic input/output system 533 (BIOS), containing the basic routines that help to transfer information between elements within computer 510, such as during start-up, is typically stored in ROM 531. RAM 532 typically contains data and/or program modules that are immediately accessible to and/or presently being operated on by processing unit 520. By way of example, and not limitation,
The computer 510 may also include other removable/non-removable, volatile/nonvolatile computer storage media. By way of example only,
The drives and their associated computer storage media discussed above and illustrated in
The computer 510 may operate in a networked environment using logical connections to one or more remote computers, such as a remote computer 580. The remote computer 580 may be a personal computer, a server, a router, a network PC, a peer device or other common network node, and typically includes many or all of the elements described above relative to the computer 510, although only a memory storage device 581 has been illustrated in
When used in a LAN networking environment, the computer 510 is connected to the LAN 571 through a network interface or adapter 570. When used in a WAN networking environment, the computer 510 typically includes a modem 572 or other means for establishing communications over the WAN 573, such as the Internet. The modem 572, which may be internal or external, may be connected to the system bus 521 via the input interface 560, or other appropriate mechanism. In a networked environment, program modules depicted relative to the computer 510, or portions thereof, may be stored in the remote memory storage device 581. By way of example, and not limitation,
The communications connections 570, 572 allow the device to communicate with other devices. The communications connections 570, 572 are an example of communication media, as discussed above.
The methods of the insurance rating system embodiments described above may be implemented in part or in their entirety using one or more computer systems such as the computer system 500 illustrated in
Some or all calculations performed in the insurance rating system embodiments described above (e.g., calculations for determining an in-home data pattern, calculations for determining an insurance premium adjustment, etc.) may be performed by a computer such as the computer 510, and more specifically may be performed by a processor such as the processing unit 520, for example. In some embodiments, some calculations may be performed by a first computer such as the computer 510 while other calculations may be performed by one or more other computers such as the remote computer 580. The calculations may be performed according to instructions that are part of a program such as the application programs 535, the application programs 545 and/or the remote application programs 585, for example.
Indicating premium adjustments (or other incentives for practicing low loss behaviors), as described in the above embodiments, may also be performed by a computer such as the computer 510. The indications may be made by setting the value of a data field stored in the ROM memory 531 and/or the RAM memory 532, for example. In some embodiments, indicating premium adjustments or other incentives may include sending data over a network such as the local area network 571 or the wide area network 573 to another computer, such as the remote computer 581. In other embodiments, indicating premium adjustments or other incentives may include sending data over a video interface such as the video interface 590 to display information on an output device such as the monitor 591 or the printer 596, for example.
