Infrastructure Management Decision-Making with Condition Data
Generated by Remote Sensors: A Time Series Framework


PROJECT COMPLETE

Research Summary Series (coming soon)

Project Report Now Available:
Click here to download the project report

Click here to view the text of the report in html


Project Number:             04-03

Research Project:           Infrastructure Management Decision-Making with Condition Data Generated by Remote Sensors: A Time Series Framework            

P.I. Name & Address:    Pablo Durango-Cohen
    Department of Civil and Environmental Engineering
    Northwestern University Transportation Center
    2145 Sheridan Road, A335 -- Evanston, IL 60208
    Phone: (847) 491-4008  Fax: (847) 491-4011
    Email: pdc@northwestern.edu  

Project Objective:          The objectives of this research study are to develop tools that will allow agencies to process and exploit the data to support infrastructure maintenance and rehabilitation (IM&R) decision-making, and to provide a framework to evaluate different strategies for deploying sensing technologies.

Project Abstract:            The researchers propose to develop an optimization model for IM&R decision-[j1] making that can accommodate potentially large quantities of condition data generated by sensors or other non-destructive evaluation technologies. The data will be processed by an algorithm known as the Kalman Filter to obtain a representation of facility condition meaningful for IM&R decision-making. The model builds on a previously presented framework that transforms condition data generated by sensors into a representation of condition that consists of two elements: structural integrity and functional performance. The motivation for the research is to capture (i) the benefits of updating/fine-tuning the representation of condition over a facility’s life-cycle in response to new data (generated by sensors), and (ii) the benefits of using different representations for different facilities. The elements used to represent the condition of steel bridges may be different than those used for concrete bridges.

Task Descriptions:    

Task 1- Literature Review – An extensive literature review will be conducted to identify possible approaches to address the research problem. The most promising is the Kalman Filter

Task 2- Model Formulation and Solution – (a) model IM&R decision-making as a latent decision process model and (b) implement Kalman Filtering algorithm

Task 3- Case Studies  – facility TBD - compare the methodology to state-of-the-art models such as the ones used in the Pontis Bridge Management System.

Task 4 – prepare reports and deliverables

Task 5 - explore the feasibility of transferring the proposed methodology into an infrastructure management system

Dates:                         Project Start Date: June 1, 2003 Project End Date: May 31, 2004

Budget:                       $77,105

Student Involvement: 1 Graduate student assistant 

Relationship to Other Research Projects:     None.

Technology Transfer Activities:          Research team will explore the feasibility of transferring the proposed methodology into an infrastructure management system and choose a state DOT partner to work with.

Potential Benefits of the Project:        New technologies require new methods to process their nascent output to a form that is meaningful and useful for deterioration modeling, and in turn, for infrastructure decision-making. This research will maximize a system’s performance by incorporating data generated by remote sensors.

TRB Keywords:         Remote Sensing, Maintenance and Rehabilitation, Infrastructure Management

Primary Subject:         Infrastructure maintenance and rehabilitation (IM&R) decision-making

Modal Orientation:     Multi-modal, primarily highway