Los Alamos National LaboratoryEngineering Institute
Addressing national needs by fostering specialized recruiting and strategic partnerships

Rapid Integrated Structural Health Assessment Using Unmanned Delivery Platform

Collaboration between Los Alamos National Laboratory and the University of California at San Diego (UCSD) Jacobs School of Engineering


  • Institute Director
  • Charles Farrar
  • (505) 663-5330
  • Email
  • UCSD EI Director
  • Michael Todd
  • (858) 534-5951
  • Executive Administrator
  • Ellie Vigil
  • (505) 667-2818
  • Email
  • Institute Administrator
  • Vacant

UCSD Faculty and Graduate Students

  • Professor Michael Todd , Structural Engineering
  • Professor Rajesh Gupta, Computer Science and Engineering
  • Professor Sanjoy Dasgupta, Computer Science and Engineering
  • Professor Tajana Rosing, Computer Science and Engineering
  • Professor Douglas Palmer, Cal it^2

LANL Collaborators

  • Dr. Matt Bement (INST-OFF: INSTITUTES)
  • Dr. Francois Hemez (X-1-MV: METHODS AND VERIFICATION)
  • Dr. Gyuhae Park (INST-OFF: INSTITUTES)
  • Dr. Chuck Farrar (INST-OFF: INSTITUTES)

The of the proposed task is to explore a novel approach to structural health monitoring that exploits the latest advances in remote structural sensing, communications, networking and data analysis. Specifically, we seek to directly address challenges related to scalability and exploitation of RFID (Radio Frequency Identification) based sensing, in situ data analysis, and model-based reasoning for rapid, economical, and reliable assessment of damage in defense-related infrastructure.  We will demonstrate the use of a remote interrogation platform based on robotic vehicles in reliable SHM using RFID-based peak-level strain sensors and advanced piezoelectric patches, sensor optimization strategies, ad hoc networking, and embedded data collection and local processing. This will advance the SHM practice from primarily visual inspection methods to quantitative analysis of various system responses and their impact on overall health through model building and model-based reasoning.
The first embedded sensing module proposed in this project will be based on the use of peak-strain sensors that can monitor the maximum strain or displacement and that can be easily integrated into an LC circuit in RFID-tagged sensor systems.  The sensing portion of the task will evolve to the use of active wireless piezoelectric strain sensors.  With an “active” sensing system, the structure in question is locally excited by a known input, and the responses are measured by the same device providing the excitation source.  Piezoelectric sensor/actuators are inexpensive (<$1 per patch) and low-power consumers (<milliwatts). The dual actuation/sensing capability allows the implementation of numerous active damage detection approaches already developed by UCSD and LANL.
The second portion of the project will expand a current software toolbox, DIAMOND II, that has been developed at LANL for damage detection.  This software contains a suite of routines for data normalization, data cleansing, data compression, data fusing, feature extraction and statistical modeling.  The user can choose the algorithms most relevant to a particular application, link them together, and develop a SHM process tailored for a particular application.  Such algorithms will include auto-associative neural networks, autoregressive/exogenous inputs data modeling, extreme value statistical modeling, efficient high-dimensional clustering, dimensionality reduction, and interdependence measures among high-dimensional data.
Finally, the third portion of the task takes monitoring with robotic vehicles to the next level, as the vehicles will monitor critical systems based upon a GPS locator and then begin to perform an inspection without human intervention.  The vehicles will search for the RFID tags on the structure and gather critical data needed to perform the system health assessment.  We will tailor the robotic vehicles to the specific defense-related applications..  A specialized controller and electronics package will be designed to accommodate both sensor modalities and RFID tag tracking.