Fault Identification and Risk Management (FIRM)
Recently, Deep Learning (DL) approaches has gained popularity as a strong tool for accurate diagnosis and prognosis of failure within structural and dynamic systems. The availability of large volumes of data along with recent developments in computational efficiency of learning algorithms, has lead the researches to investigate data-driven methods for challenging SHM application. Conventional data-driven Health Monitoring (HM) consists of three main steps: hand-crafted feature design, feature extraction/selection and model training; which makes the implementation of fault diagnosis and prognosis approaches susceptible to change in environmental conditions and less reliable in practical applications. In this project capability of deep learning to cope with large volumes of data and representing multi-domain/multi-sensory data in an appropriately compressed and reliable manner, is being investigated for development of novel emerging intelligent HM systems.
Snake robots are a class of hyper-redundant (many degrees of freedom) mechanisms that move by replicating interval shape-changes inspired by snakes and worms. Their narrow cross-section area and the extreme range-of-motion of the joints allow them to navigate many diverse environments, such as pipes, channels, uneven ground, and internal hard-to-access pipe areas. For example, the use of inspection and rescue robots played an important role in emergency-response to the nuclear accident at Fukushima Daiichi nuclear power plants. Particularly, the harsh and hazardous environments of nuclear applications require a reliable and resilient robot. Snake robots are well-suited to pipe inspection deployment due to their ability to actively locomote in a wide range of pipe diameters and configurations with a single mechanism. In this project, novel bio-inspired robotic crawlers is being developed for sea systems and nuclear submarine piping inspection and repair. While the majority of research in pipe crawlers has been primarily focused on obtaining access and providing visual feedback, there are many important challenges that still remain unaddressed, particularly in the pipeline integrity monitoring and repair technologies. Two of these limitations that will be overcome in the current research are 1) the lack of sufficient payload capability for carrying inspection and repair equipment 2) the inability of current Non-Destructive Testing and Evaluation (NDT&E) methods to cope with in-field inaccuracies of measurements and strict nuclear environment deployment requirements.
Robotic Autonomous Inspection Systems Education (RAISE)
The Robotic Autonomous Inspection Systems Education (RAISE) program seeks STEM workforce development through a training curriculum that incorporates supervised research opportunities in robotics with SDSMT faculty advisors with close collaborations with federal agencies and private sector. The program will recruit multiple new fellows per year (graduate and undergraduate students) and build on the existing core project-based design curricula of SDSMT while introducing specialized robotics topics . The research opportunity aims to produce specialists who have the skills required to envision, develop, and deploy autonomous and semi-autonomous systems that directly address intelligent maintenance mission. The RAISE training program in robotics is formulated to provide a course of study that balances theory and practice with opportunities for field deployment and evaluation in partnership with sponsoring partners toward developing a future workforce that has experience in solving field challenges through advancement in robotic technologies.