Asset-level vulnerability and mitigation
Methods, systems, and apparatus for receiving a request for a damage propensity score for a parcel, receiving imaging data for the parcel, wherein the imaging data comprises street-view imaging data of the parcel, extracting, by a machine-learned model including multiple classifiers, characteristics of vulnerability features for the parcel from the imaging data, determining, by the machine-learned model and from the characteristics of the vulnerability features, a damage propensity score for the parcel, and providing a representation of the damage propensity score for display.
1. A computer-implemented method comprising:
receiving a request for a damage propensity score for a parcel for one or more hazard event scenarios;
receiving imaging data for the parcel, the imaging data capturing an aspect of the parcel;
extracting, by a trained machine-learned model comprising a plurality of classifiers, characteristics of a plurality of vulnerability features for the parcel from the imaging data;
determining, by the trained machine-learned model and from the characteristics of the plurality of vulnerability features, the damage propensity score for the parcel for the one or more hazard event scenarios;
determining, by the trained machine-learned model and from the characteristics of the plurality of vulnerability features for the parcel, a plurality of mitigation steps for reducing the damage propensity score for the one or more hazard event scenarios;
selecting, by the trained machine-learned model and from the plurality of mitigation steps, a proposed subset of one or more mitigation steps, wherein the selecting comprises,
for each subset of one or more mitigation steps:
determining, by the trained machine-learned model and based on the selected subset of one or more mitigation steps, a corresponding updated damage propensity score; and
selecting the proposed subset of one or more mitigation steps, wherein the proposed subset of one or more mitigation steps corresponds to the updated damage propensity score yielding a target reduction in the damage propensity score for the one or more hazard event scenarios; and
providing a representation of the subset of one or more proposed mitigation steps and the corresponding updated damage propensity score for display.
2. The computer-implemented method of claim 1 , wherein the imaging data comprises street-view imaging data of the parcel.
3. The computer-implemented method of claim 2 , wherein the imaging data further comprises one or more of LIDAR data, high-resolution satellite image data, aerial image data, infrared image data, and user-provided images.
4. The computer-implemented method of claim 1 , wherein the target reduction comprises a largest reduction between the updated damage propensity score and the damage propensity score for the subsets of one or more mitigation steps.
5. The computer-implemented method of claim 1 , wherein selecting the proposed subset of one or more mitigation steps comprises selecting the proposed subset of one or more mitigation steps based in part on a cost of implementing the proposed subset of one or more mitigation steps and a resulting reduction of the updated damage propensity score from the damage propensity score.
6. The computer-implemented method of claim 1 , wherein a first proposed subset of one or more mitigation steps for a first hazard scenario is different than a second subset proposed of one or more mitigation steps for a second hazard scenario.
7. The computer-implemented method of claim 1 , wherein extracting the plurality of mitigation steps comprises:
identifying, by the trained machine-learned model, adjustments to the characteristics of the plurality of vulnerability features extracted from the imaging data, and
wherein determining the updated damage propensity score comprises determining the updated damage propensity score based on the adjustments to the characteristics of the plurality of vulnerability features.
8. The computer-implemented method of claim 1 , further comprising:
selecting, for each of at least two hazard event scenarios, a corresponding proposed subset of one or more mitigation steps for each of the at least two hazard event scenarios; and
providing a representation of the corresponding proposed subsets for each of the at least two hazard event scenarios for display.
9. The computer-implemented method of claim 1 , wherein selecting each subset of one or more mitigation steps from the plurality of mitigation steps comprises:
selecting at least two different subsets of one or more mitigation steps from the plurality of mitigation steps.
10. A non-transitory computer storage medium encoded with a computer program, the computer program comprising instructions that when executed by a data processing apparatus cause the data processing apparatus to perform operations comprising:
receiving a request for a damage propensity score for a parcel for one or more hazard event scenarios;
receiving imaging data for the parcel, wherein the imaging data comprises an aspect of the parcel;
extracting, by a trained machine-learned model comprising a plurality of classifiers, characteristics of a plurality of vulnerability features for the parcel from the imaging data;
determining, by the trained machine-learned model and from the characteristics of the plurality of vulnerability features, the damage propensity score for the parcel for the one or more hazard event scenarios;
determining, by the trained machine-learned model and from the characteristics of the plurality of vulnerability features for the parcel, a plurality of mitigation steps for reducing the damage propensity score for the one or more hazard event scenarios;
selecting, by the trained machine-learned model and from the plurality of mitigation steps, a proposed subset of one or more mitigation steps, wherein the selecting comprises,
for each subset of one or more mitigation steps:
determining, by the trained machine-learned model and based on the selected subset of one or more mitigation steps, a corresponding updated damage propensity score; and
selecting the proposed subset of one or more mitigation steps, wherein the proposed subset of one or more mitigation steps corresponds to the updated damage propensity score yielding a target reduction in the damage propensity score for the one or more hazard event scenarios; and
providing a representation of the subset of one or more proposed mitigation steps and the corresponding updated damage propensity score for display.
11. The non-transitory computer storage medium of claim 10 , wherein the target reduction comprises a largest reduction between the updated damage propensity score and the damage propensity score for the subsets of one or more mitigation steps.
12. The non-transitory computer storage medium of claim 10 , wherein selecting the proposed subset of one or more mitigation steps comprises selecting the proposed subset of one or more mitigation steps based in part on a cost of implementing the proposed subset of one or more mitigation steps and a resulting reduction of the updated damage propensity score from the damage propensity score.
13. The non-transitory computer storage medium of claim 10 , wherein a first proposed subset of one or more mitigation steps for a first hazard scenario is different than a second subset proposed of one or more mitigation steps for a second hazard scenario.
14. The non-transitory computer storage medium of claim 10 , wherein extracting the plurality of mitigation steps comprises:
identifying, by the trained machine-learned model, adjustments to the characteristics of the plurality of vulnerability features extracted from the imaging data, and
wherein determining the updated damage propensity score comprises determining the updated damage propensity score based on the adjustments to the characteristics of the plurality of vulnerability features.
15. The non-transitory computer storage medium of claim 10 , further comprising:
selecting, for each of at least two hazard event scenarios, a corresponding proposed subset of one or more mitigation steps for each of the at least two hazard event scenarios; and
providing a representation of the corresponding proposed subsets for each of the at least two hazard event scenarios for display.
16. The non-transitory computer storage medium of claim 10 , wherein selecting each subset of one or more mitigation steps from the plurality of mitigation steps comprises:
selecting at least two different subsets of one or more mitigation steps from the plurality of mitigation steps.
17. A system comprising:
a user device; and
one or more computers operable to interact with the user device and to perform operations comprising:
receiving a request for a damage propensity score for a parcel for one or more hazard event scenarios;
receiving imaging data for the parcel, wherein the imaging data comprises an aspect of the parcel;
extracting, by a trained machine-learned model comprising a plurality of classifiers, characteristics of a plurality of vulnerability features for the parcel from the imaging data;
determining, by the trained machine-learned model and from the characteristics of the plurality of vulnerability features, the damage propensity score for the parcel for the one or more hazard event scenarios;
determining, by the trained machine-learned model and from the characteristics of the plurality of vulnerability features for the parcel, a plurality of mitigation steps for reducing the damage propensity score for the one or more hazard event scenarios;
selecting, by the trained machine-learned model and from the plurality of mitigation steps, a proposed subset of one or more mitigation steps, wherein the selecting comprises,
for each subset of one or more mitigation steps:
determining, by the trained machine-learned model and based on the selected subset of one or more mitigation steps, a corresponding updated damage propensity score; and
selecting the proposed subset of one or more mitigation steps, wherein the proposed subset of one or more mitigation steps corresponds to the updated damage propensity score yielding a target reduction in the damage propensity score for the one or more hazard event scenarios; and
providing a representation of the subset of one or more proposed mitigation steps and the corresponding updated damage propensity score for display.
18. The system of claim 17 , wherein the target reduction comprises a largest reduction between the updated damage propensity score and the damage propensity score for the subsets of one or more mitigation steps.
19. The system of claim 17 , wherein selecting the proposed subset of one or more mitigation steps comprises selecting the proposed subset of one or more mitigation steps based in part on a cost of implementing the proposed subset of one or more mitigation steps and a resulting reduction of the updated damage propensity score from the damage propensity score.