PITA Fiscal Year 2010 Projects - Innovative Infrastructure Systems

Feasibility Study of a Bus-Data-Driven Approach for Assessing Damage in Bridge Structures Using Multiresolution Signal Processing Techniques

Principal Investigators: James H. Garrett, Jacobo Bielak, Christoph Mertz

This project aims at using pseudo data from idealized vehicle-bridge systems, available data from a moving bus, and data acquired in the future from a new vehicle and bridge, to study the feasibility of a new approach for bridge structural health monitoring. The approach has the potential of causing a transformative change in the way bridges are monitored.

Today, structural evaluation is based primarily on subjective, biennial visual inspections. There are approximately 600,000 highway bridges in the U.S., 25% of which the Federal Highway Administration in 2007 rated as structurally deficient or functionally obsolete. The initial investment cost of monitoring bridges by placing sensors directly into the structure, is estimated at roughly $22.5 billion for the structurally deficient bridges, and $90 billion for all 600,000. The indirect monitoring approach, if successful, would reduce this cost drastically. This fleet-based approach will allow for much broader coverage of the entire bridge population, where only a fraction of the bridges are sensed directly today due to initial and long-term maintenance costs. Data will be processed by image processing experts from the Department of Biomedical Engineering; it is a civil engineering application that requires expertise in structural dynamics and numerical modeling; and it is highly technological in terms of sensing equipment required for data acquisition, storage, and management. These last tasks will be assisted by the NAVLAB (The Robotics Institute) members of the team. For future large-scale applications we have already been offered the use of necessary computational resources by the Pittsburgh Supercomputing Center (PSC). One of the Co-PIs is already a regular user of the PSC facilities. Initial research, in which the response analysis of a simple oscillator-beam interaction model was examined, has produced preliminary results that show the potential of the approach.This project is being proposed to be funded by two different sources: Traffic21 and PITA. We are asking for about $152,000 in direct costs from Traffic21 and about $23,000 in direct costs from PITA.