Decision Support System

Data/information flow and decision support: The purpose of this system is, rather than overwhelm them in raw data, to present remote scour information to bridge owners in an easy to use, easy to understand format to aid in decision making. Coordination of the flow of data and information between the remote sensing system and the decision support client will be a major challenge. The bulk of data processing is embedded in the sensors themselves. The actionable and decision-enabling outputs that are produced from sensor hardware are: flow or no-flow (static versus dynamic) state of the sensor (or a sensor fault condition detected), the channel bed profile, and the time-series of scour depth over a user-defined time period. In addition, user-defined parameters and thresholds will contribute even more to a rich, decision-making workflow. An example of added-value processes and products is the provision of critical scour depth alerts (e.g. an e-mail or application pop-up to notify decision makers about a recent critical scour depth measurement) based on user-defined thresholds. Another example are report templates, stored on the server, dynamically populated with up-to-date bridge information, and generated quickly upon request in a variety of formats (e.g. Microsoft Word Document or Adobe PDF).


Sample 2D scour dept visualizations for a set of bridge piers

The project team will work closely with its cost-share partners, the Michigan Departments of Transportation and Maryland State Highway Administration (MDSHA) as well as the TAC and survey respondents to develop visualization and analysis routines that best support decision makers, and establish sample intervals and alarm conditions. Flow or no-flow states can be visualized a variety of ways, such as the following figure, but some visualizations may be more meaningful than others. The project team will assess, for instance, whether visualizing the scour depth in three-dimensions around each bridge pier is necessary or useful for the decision maker or, as an alternative, it is more appropriate to view the average scour depth for each pier.

Modification to existing Bridge Condition DSS visualize to show scour information

Software support framework for DSS engine

Provision of remote access is another important challenge. The Bridge Condition Decision Support System (DSS), developed under an existing contract with USDOT-RITA, already has the capabilities needed to support the visualizations described shown in the first succeeding figure. Developed entirely with open-source software, the DSS exposes bridge data for decision support through a series of web services hosted on an Apache server latter figure. The web services are programmed in Python using the Django web framework, a database application programming interface (API) that allows (among other things) for database queries to be parameterized and executed programmatically. The parameterization of queries is done in the application layer, on the client, by decision makers who, for instance, wish to look at the first 50 bridge records in the inventory sorted by structure number. Such a request is handled by the web server which delegates Django to construct a database query and return the desired results, encoded in Javascript Object Notation (JSON, a more compact alternative to XML), through one of the web services. The client application responsible for making these requests is a Javascript program, utilizing the ExtJS framework, which contains all the logic necessary to provide the decision-maker with an informative, user-friendly experience allowing data manipulation all on its own until more data are needed. The application also retains state; it remembers what the user has done in the same session so that, after filtering a table of bridge records, a subsequent sort operation returns only the filtered records. 

To support potentially sophisticated structural views of bridge piers, channel profile views and other visualizations, the DSS might be modified by the addition of a graphics library on the server. This library would be used to generate pre-rendered decision-support visualizations based on up-to-date scour data. This is in contrast to the client-side graphics library for charting and plotting, already implemented in the DSS. In addition, temporal analysis routines stored on the server can be run to generate value-added bridge scour decision support products such as rate of scour and time to a threshold scour depth.