What it does

FaultLine is an integrated application that broadcasts an alert from a central system which monitors real time occurrences of earthquakes. An index of risk (i) is calculated using weighted environmental and spatial factors and assigned to areas of a building and given the threshold of the risk, the area is classified as dangerous or safe. During an earthquake, the received message is evaluated by user's device to inform the user of what actions to take based on their position in the building.

How I built it

  1. a) Creation of Reaction Information for Facilities Tectonics (RIFT) using Geo processing and Business Analyst tools in ArcGIS Pro.

RIFT is a workflow created to integrate CAD data structured for the ArcGIS indoors application into a Polygon Feature Class which depicts the suitability of response action during an earthquake.

Data Source: CAD Data

b) Spatial Analysis for Weighted Suitability.

Suitability Analysis allows you to qualify, compare, and rank candidate sites based on how closely they adhere to criteria that you select and define.

  • The CAD data set for buildings contains important information concerning infrastructure, number of office spaces, the use of spaces, objects, exits, and safety features. This information is stored as Feature layers containing polygon, line and point geometries. Suitability Analysis is performed on polygonal inputs that are formatted as the Suitability Analysis layer. Our suitability layer is the Polygon Feature Class of the Building "L" on Esri Redlands campus.
  • Each building layer contains polygons for delineating each area and is assigned a unique code i.d, this is the polygon layer which is populated with weighted values which is used to determining the suitability of an area.
  • We identified ten variables to be used in the construction of the Weighted Suitability Analysis and ranked them from 1-10 based on their influence of threat. The minimum distance to an emergency exit: Min Dis The number of elevators present: ElevatorCount The Minimum distance from stairway: Stair_Min_Dis The Cubicle count: Cubcile_count The ATM count: ATM_count The Vending Machine Count: Vending_count Doorway Count: Doorway_count Column Count: Col_count Glass Walls Count: GWall_count Glass Sliding Doors: Sliding_count

The analyses of the suitability layer candidates are ranked and scored based on these weighted variables. We used the tool: Add point layer based suitability criteria to populate the Suitability layer polygons with weighted values. This tool adds our criteria based on spatial relationships between the input layer and a given point layer. E.g. The Safety and Security Point Layer with the CAD Data creates a feature attribute in our suitability layer which is the minimum distance of that polygon from an emergency exit. As the distance increases the suitability of that polygon as an area to be evacuated decreases, due to that, the presence of emergency exits are more important to the suitability than,lets say an ATM machine, this distance value is weighted more than the presence of an ATM. Therefor as the distance increases it has a stronger effect on the classification than an increase in the count of the ATM machine presence.

Some of the feature layers provided within the CAD data set are actually polygon and line features. In order to integrate these, such as the variable "Glass Sliding Doors" it is necessary to extract point features. Using the line to points tool, points are generated along the line for these features, and then used as the point information for populating those weighted variables in the suitability layer.

-A bi-variate suitability score is calculated by comparing the criteria across all candidate sites. Each criterion receives a weighted score, which are returned as new attributes. A final score is also returned. It summarizes individual weighted scores into an overall ranking. Once the variables were ranked, we assigned percentiles of influence using the Calculate Suitability Score geo-processing tool. With further research, these percentiles and ranks can be further refined. However for our purposes we needed an increasing amount of influences with each successive ranking of the variables. Xi (Suitability Score)= Weighted SUM(x1, x2, x3....x10) Sum(1 + 2 + 3....10) = 55 X1= 1/55 * 100 = 1.818182% X2= 2/55 * 100 = 3.636364% .... X10= 10/55 * 100 = 18.18182%

After running the Calculate Suitability Score, classes of suitability are assigned values based on the weighted criteria, we sorted these into two different quantiles. For the beta we used an equal interval symbology, essentially creating two different classes. The first class, with the lowest suitability score, is actually the class that determines that a user is in a safe location. The threshold for classification of the range has been arbitrarily set in this case. However using more variables and a determined index, the polygons will be classified as they fall on either side of the range.

The result is a suitability analysis layer for a building, with each feature assigned a suitability score, and sorted into one of two classes.

  1. Feature publishing online.

The suitability layer is copied using the Copy Features tool and shared as a feature package to ArcGIS online.

  1. The mobile application loads the feature layer for the building and allows users to register their offices for alerts. We then use this information to evaluate the proper message to show a user during an earthquake.

We created a server that constantly listens for earthquake alerts from USGS using their web live feed endpoint. Upon receipt of the notification, our server sends a signal to all the devices subscribed to the service in realtime. When a device receives the signal, it looks at the stability score of the user's space and provides an alert on screen indicating what the best action is.

We built the mobile app for android with Java. We built the server using Firebase's realtime databases and cloud functions in order to provide realtime notifications to users.

  1. Device Listens
  2. Device Sends Notification

Challenges

  1. Indicating Weighted Suitability and Ranking- The information to make structured and efficient decisions for ranking the variables was available, however an index of danger needs to be generate to classify the building regions based on their suitability. We used a quantile classification containing two classes. However, when moving forward from Beta a threshold index will be determined, where the final score of suitability falls above, or below this threshold, determines the action to be taken. Creating a structured and educated way of ranking the variables for the beta, set the foundations for moving forward, however standardizing this was the challenge (See "How we did it" for reference to ranking and suitability scoring.

Accomplishments that I'm proud of

Helping people stay safe during earthquakes and aiding in assisting the foundation for further technological applications for earthquake response.

What's next for FaultLine

-Integrating of SDK through ArcGIS Indoors for more precise and realtime evaluation of risk area and message reposes. -Deploying to different platforms such as desktop and iOS. *CITATIONS*

  1. Smith, Keith. Environmental hazards: assessing risk and reducing disaster. Routledge, 2003. Accessed July 27th, 2019
  2. Mileti, Dennis S., and Paul W. O'Brien. "Warnings during disaster: Normalizing communicated risk." Social Problems 39.1 (1992): 40-57. Accessed July 25th, 2019
  3. Sorensen, John H. "When shall we leave? Factors affecting the timing of evacuation departures." International journal of mass emergencies and disasters 9.2 (1991): 153-165. Accessed July 27th, 2019
  4. https://earthquake.usgs.gov/earthquakes/eventpage/ci38457511/executive

Built With

Share this project:
×

Updates