US20160283874A1

US20160283874A1 - Failure modeling by incorporation of terrestrial conditions

Info

Publication number: US20160283874A1
Application number: US14/665,247
Authority: US
Inventors: Aanchal Aggarwal; Younghun Kim; Tarun Kumar; Abhishek Raman
Original assignee: International Business Machines Corp
Current assignee: Utopus Insights Inc
Priority date: 2015-03-23
Filing date: 2015-03-23
Publication date: 2016-09-29
Also published as: US20160283915A1; US12211008B2

Abstract

A system and method to schedule inspections for a plurality of assets at a corresponding plurality of outdoor locations are described. The method includes retrieving location attributes of each of the plurality of assets, retrieving terrestrial geotagged images that include one or more of the plurality of assets, obtaining parameters associated with each of the plurality of assets based on information from the images, and determining risk factors associated with each of the plurality of assets based on the associated parameters and a failure mapping. The method also includes determining a risk score for each of the plurality of assets based on the corresponding risk factors, and scheduling inspections of the plurality of assets based on the risk score associated with each of the plurality of assets.

Description

BACKGROUND

The present invention relates to management of assets, and more specifically, to failure modeling by incorporation of terrestrial conditions.
In many applications, physical assets are located in various outdoor environments. Utility poles, windmills, water towers, and cellular towers are such exemplary assets. The number of these assets makes regular inspection of their condition impractical. Yet, regular inspection and maintenance are important to the proper functioning of the overall system.

SUMMARY

According to one embodiment of the present invention, a method of scheduling inspections for a plurality of assets at a corresponding plurality of outdoor locations includes retrieving location attributes of each of the plurality of assets; retrieving terrestrial geotagged images that include one or more of the plurality of assets; obtaining parameters associated with each of the plurality of assets based on information from the images; determining risk factors associated with each of the plurality of assets based on the associated parameters and a failure mapping; determining a risk score for each of the plurality of assets based on the corresponding risk factors; and scheduling inspections of the plurality of assets based on the risk score associated with each of the plurality of assets.
According to another embodiment, a system to schedule inspection of a plurality of assets at a corresponding plurality of outdoor locations includes an input interface configured to obtain a selection of the plurality of assets; a processor configured to obtain terrestrial geotagged images that include one or more of the plurality of assets, obtain parameters associated with each of the plurality of assets based on information from the images, determine risk factors associated with each of the plurality of assets based on the associated parameters and a failure mapping, and determine a risk score for each of the plurality of assets based on the corresponding risk factors; and an output interface configured to output an inspection schedule for the plurality of assets based on their corresponding risk score.
Additional features and advantages are realized through the techniques of the present invention. Other embodiments and aspects of the invention are described in detail herein and are considered a part of the claimed invention. For a better understanding of the invention with the advantages and the features, refer to the description and to the drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

The subject matter which is regarded as the invention is particularly pointed out and distinctly claimed in the claims at the conclusion of the specification. The forgoing and other features, and advantages of the invention are apparent from the following detailed description taken in conjunction with the accompanying drawings in which:

FIG. 1 is a process flow of a method of scheduling inspections of assets according to embodiments of the invention;

FIG. 2 is a process flow of a method of obtaining parameter values according to embodiments of the invention;

FIG. 3 illustrates an image from which a set of parameter values are obtained according to embodiments of the invention;

FIG. 4 illustrates an image from which another set of parameter values are obtained according to embodiments of the invention; and

FIG. 5 is a block diagram of an exemplary system to schedule inspections of assets according to embodiments of the invention.

DETAILED DESCRIPTION

As noted above, outdoor assets of a system require inspection and maintenance to ensure their proper condition and, in turn, the proper operation of the overall system. However, the number of these assets can make regular inspections challenging. Further, while some known attributes, such as installation date and material type, may suggest a given inspection schedule, environmental conditions (e.g., sunlight exposure, soil moisture level, ground relative temperature) can dramatically change the frequency with which inspection and maintenance must be performed. Embodiments of the systems and methods detailed herein relate to scheduling inspection of outdoor assets based on mapping variables, at least some of which are obtained via image processing, to failure probability.
FIG. 1 is a process flow of a method of scheduling inspections of assets according to embodiments of the invention. At block 110, risk factors are obtained based on asset inspections and quantified to obtain a risk vector R for each asset of a set of inspected assets. Exemplary risk factors include mechanical breakdown, rotting progress (of a wooden pole, for example), animal invasion, treatment aging, vine crawling, angle (e.g., whether the asset is tilted or straight), and the risk of a tree falling. To obtain the risk vector R for each asset, a numerical score is assigned to a risk factor according to a predetermined range for each factor. Each risk factor is a continuous value contributing to a final risk score ranging 0 to 1 with 1 being the highest risk. Each risk factor may be calculated based on a risk feature in the image. For example, the total number of pixels from a forest area divided by the total number of pixels in a single image may be one risk factor, a forest risk factor. An average distance (in linear length) to the intersection areas in the map may be another risk factor. The final risk score is a weighted linear combination of all the risk factors. At block 120, parameters are obtained for the inspected assets. An exemplary parameter includes the ratio, within a specified area that includes the asset, among water, barren land, grassy area, wooded area, roads and buildings. Other exemplary parameters include the relative location of an asset to the location of trees or high-rise buildings, distance between an asset and the closest road and intersection, size and traffic conditions of the nearest road, and population of proxy for road congestion in a specified region around the asset. Each of these exemplary indications is parameterized, as further discussed below, to provide a parameter vector P associated with each asset. These parameters need not be obtained via inspection. For inspected assets and other assets (per block 150), the parameters may be obtained through geotagged image data (e.g., satellite images), for example. The process of obtaining the parameter vector P via image data is further detailed below.
At block 130, the risk factors and parameters associated with inspected assets are used to determine a failure mapping. In machine learning, the process described below of determining the failure mapping matrix F may be referred to as learning. Based on:
R ^k =FP ^k +N ^k [EQ. 1]
For each inspected asset k, the risk factors discussed above are expressed as risk vector R^k, the parameters discussed above are expressed as the parameter vector P^k, and error factors are expressed as error vector N^k. Exemplary error factors are associated with variability, parameterization error, and clerical error. EQ. 1 may be used to solve for the linear failure mapping matrix F by using known techniques to minimize the error vector N such that each F vector (associated with each inspected asset) is solved by minimizing:
|R−FP| _p [EQ. 2]
The norm of the matrix is indicated by p. The risk vector is subject to:
0≦R≦1 [EQ. 3]
R=[R ¹ ,R ² , . . . R ^h]^T [EQ. 4]
The number of observations is given by h. Also,
P=[P ¹ ,P ² , . . . P ^h]^T [EQ. 5]
The result provides
F=[F,F, . . . F] ^T [EQ. 6]
At block 140, the failure mapping matrix (F) solved using EQ. 2 on the inspected assets is employed to determine risk factor vector R for each of the uninspected assets that are not inspected. At block 150, parameters are ascertained for uninspected assets from geotagged images (e.g., satellite images) as further described below. The parameter vectors P of each of the uninspected assets are used in EQ. 1, along with the failure mapping matrix F, which was obtained by solving EQ. 2 at block 130, to determine the risk vectors R. Parameter values (obtained from the images at block 150) are assigned risk points at block 160. The risk vector R obtained for each uninspected asset (at block 140), in addition to the risk points assigned to the parameters (at block 160) are combined to determine risk scores for each of the uninspected assets at block 170. At block 180, scheduling inspections is based on the risk scores determined at block 170. For example, a wooden electric pole may have a list of risk factors associated with the normal distance to cross-section, the electric pole being in wooded area, and near vines, which will have a risk vector of [0.1 0.8 0.9]. This indicates a low risk (0.1) due to the cross-section, because the cross-section is far, but a high risk (0.8, 0.9) associated with rot and vine invasion. Another example could be a distribution transformer in a crowded region where each distribution transformer needs to supply many more customers resulting in high risk score due to the population score.
FIG. 2 is a process flow of a method obtaining and using parameter values according to embodiments of the invention. The parameter values may be obtained for blocks 120 and 150 (FIG. 1). In alternate embodiments, the parameters used at block 120 may be obtained via physical inspection of the area in which an asset is located. The description below is exemplary and does not represent an exhaustive list of parameters. Additional parameter or risk factors (discussed with reference to blocks 110 and 140 at FIG. 1) may be added based on the specific asset (e.g., windmill versus utility poll), for example. At block 210, selecting a set of assets may include selecting uninspected assets for which inspections must be scheduled (this pertains to block 150, FIG. 1). Selecting the set of asserts may instead include selecting inspected assets that will be used to determine the failure mapping matrix F (this pertains to block 120, FIG. 1). At block 220, retrieving location attributes of each asset selected at block 210 may include retrieving location information that was stored at the time of installation of the asset. This information may be alternately or additionally obtained from a global positioning system (GPS) associated with each asset, for example. Traditional triangulation techniques may also be used to obtain location information periodically and separate from inspections. The location attributes of each asset are used to retrieve terrestrial geotagged images around each asset location at block 230. The images may be obtained from a geographic information system (GIS) application, for example. The images may be satellite images or other images (e.g., obtained with a drone) that illustrate the landscape around one or more assets. Exemplary images are discussed below and show that parameters for more than one asset may be determined based on the same image. Information obtained from the images for each of the assets is parameterized as also discussed below.
FIG. 3 illustrates an image 300 from which a set of parameter values are obtained according to embodiments of the invention. In the image shown in FIG. 3, the assets are utility poles 310. Areas with trees are indicated as 310, and grassy areas are indicated as 320. Other types of exemplary areas that may be identified based on the contents of an image are areas with water and areas with buildings or residences. The areas may be discerned from the images in any known way. For example, each pixel of the image may be assigned a grayscale value, and each of the types of areas (e.g., trees, barren, grass) may be identified based on a grayscale range. Once the different areas within the image 300 are identified, the information may be parameterized as a ratio of the number of pixels associated with each type of area to the total number of pixels. For example:
$\begin{matrix} P_{barren} = \frac{number_of_pixels_identified_as_barren_land}{total_number_of_pixels} & [EQ . 7] \\ P_{wooded} = \frac{number_of_pixels_identified_as_wooded_land}{total_number_of_pixels} \end{matrix}$
The parameters (used at block 150, FIG. 1) also have risk points associated with them (block 160, FIG. 1). These assignments of risk points may be based on a lookup table, for example, or another predetermined association between some or all of the parameterized information and a set of risk points. For example, weather exposure (e.g., relative direct sun exposure) of an asset indicated by the image 300 may be associated with risk points. As discussed with reference to FIG. 4 below, risk points may also be computed.
FIG. 4 illustrates an image 400 from which another set of parameter values are obtained according to embodiments of the invention. Again, a utility pole 310 is indicated as an exemplary asset in the image 400. The distance 410 to the closest road and the distance 420 to the closest intersection are indicated. These distances 410, 420 may be parameterized as Euclidean distances using latitude and longitude. Given a Euclidean distance P_distance, the associated risk points may be determined as follows:
$\begin{matrix} P_{distance_risk} = \sum_{i = 1}^{k} α_{i} \frac{1}{P_{disance}^{i}} & [EQ . 9] \end{matrix}$
The number of road segments is k, and the associated risk factors for each type of the road segments is α, and P_distanceis associated with a combination of all k of the road segments.
FIG. 5 is a block diagram of an exemplary system 500 to schedule inspections of assets according to embodiments of the invention. The system 500 includes one or more processors 510 to process the information needed to assign a risk score to each asset and thereby determine a schedule of inspections. The processor 510 executes instructions stored in one or more memory devices 520. The system 500 receives information via an input interface 530. For example, location attributes of the assets and the images needed to determine parameters for each asset may be obtained by the system 500 via the input interface 530. Location information may be stored in one or more memory devices 520 in alternate embodiments. The input interface 530 may include a keyboard or other user input device as well as an interface to other processors. The input interface 530 may facilitate selection of the set of assets whose inspection schedule is to be determined. Information produced by the system 500, such as the inspection schedule, for example, is output via an output interface 540. Communication at the input interface 530 and output interface 540 may be wireless or through other known methods.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and/or “comprising,” when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one more other features, integers, steps, operations, element components, and/or groups thereof.
The corresponding structures, materials, acts, and equivalents of all means or step plus function elements in the claims below are intended to include any structure, material, or act for performing the function in combination with other claimed elements as specifically claimed. The description of the present invention has been presented for purposes of illustration and description, but is not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the invention. The embodiment was chosen and described in order to best explain the principles of the invention and the practical application, and to enable others of ordinary skill in the art to understand the invention for various embodiments with various modifications as are suited to the particular use contemplated.
The flow diagrams depicted herein are just one example. There may be many variations to this diagram or the steps (or operations) described therein without departing from the spirit of the invention. For instance, the steps may be performed in a differing order or steps may be added, deleted or modified. All of these variations are considered a part of the claimed invention.
While the preferred embodiment to the invention had been described, it will be understood that those skilled in the art, both now and in the future, may make various improvements and enhancements which fall within the scope of the claims which follow. These claims should be construed to maintain the proper protection for the invention first described.
The descriptions of the various embodiments of the present invention have been presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method of scheduling inspections for a plurality of assets at a corresponding plurality of outdoor locations, the method comprising:

retrieving location attributes of each of the plurality of assets;

retrieving terrestrial geotagged images that include one or more of the plurality of assets;

obtaining parameters associated with each of the plurality of assets based on information from the images;

determining risk factors associated with each of the plurality of assets based on the associated parameters and a failure mapping;

determining a risk score for each of the plurality of assets based on the corresponding risk factors; and

scheduling inspections of the plurality of assets based on the risk score associated with each of the plurality of assets.

2. The method according to claim 1, wherein the retrieving the terrestrial geotagged images includes retrieving satellite images.

3. The method according to claim 1, wherein the retrieving the terrestrial geotagged images includes retrieving images from a geographic information system (GIS) application.

4. The method according to claim 1, wherein the obtaining parameters based on the information from the images includes parameterizing a distance from each of the plurality of assets to a corresponding nearest road and a distance from each of the plurality of assets to a set of intersections.

5. The method according to claim 4, wherein the parameterizing the distance from each of the plurality of assets to the corresponding nearest road includes using a Euclidean distance based on latitude and longitude.

6. The method according to claim 1, wherein the obtaining parameters based on the information from the images includes parameterizing a ratio of a size of an area comprising woods, water, or grass to a total size of an area of the corresponding image.

7. The method according to claim 6, wherein the parameterizing the ratio for each area type includes solving:

P_{area_type} = \frac{number_of_pixels_identified_as_area_type}{total_number_of_pixels},

the area type being the woods, the water, or the grass.

8. The method according to claim 1, further comprising determining risk points associated with each of the plurality of assets based on the associated parameters.

9. The method according to claim 8, wherein the determining the risk score of each of the plurality of assets is based on the corresponding risk factors and the corresponding risk points of each of the plurality of assets.

10. The method according to claim 1, further comprising determining the failure mapping (F) based on inspecting and analyzing a set of assets.

11. The method according to claim 10, wherein the inspecting and analyzing the set of assets includes determining risk factors of the set of assets (R) and parameters of the set of assets (P) and solving for the failure mapping F by minimizing

|R−FP| _p

wherein p indicates the norm.

12-20. (canceled)