Prognostics and Health Management of Complex Systems

/
Prognostics and Health Management of Complex Systems

The confluence of recent advancements in sensor technology, information technology, probabilistic inference methods, physics of failure sciences, and complex system modeling offers the foundation for the next major push in realizing on-line, real-time condition monitoring, diagnostics, and prognostics for complex hybrid systems. The strong momentum in this direction is evidenced by the formation of research groups in prognostics and health management (PHM), creation of professional societies, and active interest and initiatives by various industries as well as the public sector. Other domains of interest and natural extensions are, “fault-adaptive controls”, and “methods for identifying and analyzing failure precursors”. Broad topics of research pursued at the Garrick Institute include (1) large scale complex system PHM technology development, and (2) model-based integration of diagnostics and prognostics for reliability and safety.

Read More

Resilience Engineering

/
Resilience Engineering

Major accidents and failures with high consequence are often hard to recover from, and the recovery is usually an afterthought, reactive in nature, and unanticipated by system designers. Over the past few years, the idea of making engineered systems resilient to failures, with built-in ability to recover, has gained some momentum. While the extent to which “resilience engineering” can be realized is not known, the appeal of the idea, enormous implications, and the vast research horizon are not difficult to imagine. The Garrick Institute seeks research opportunities to introduce resilience through new design paradigms and full integration of advances in relevant disciplines such as materials science, robotics, electrical engineering and computer science, and human factors.

Read More

X-Ware Systems Reliability

/
X-Ware Systems Reliability

Modern systems are a hybrid of hardware, software, and human elements or subsystem, with highly complex interactions and interdependencies (X-Ware). Never before in the history of engineered systems has the challenge of identifying and safeguarding against system vulnerabilities to failure been so formidable. While the need is clear, current reliability and safety engineering methods and tools are totally inadequate, as they tend to focus on either the hardware, or the software, or human function, and not designed to look at failures emerging from the complexity of interactions and interfaces of X-ware systems. A comprehensive approach requires methods for identifying interaction failure modes, and new design operational concepts to prevent or mitigate their effects. CRRE aims to develop the necessary methods and technologies to address this critical need.

Read More

Quantum Computing

/
Quantum Computing

The increasing size and interconnectivity of complex engineering systems as well as the global trend towards industry 4.0 and the immense amount of information demands the exploration of innovative approaches to store, maintain and process information at an unprecedented scale. Quantum computing emerges as a promising avenue to achieve diagnostics and prognostics solutions that are exponentially faster that those existing today. The center aims to leverage quantum computing to improve the risk and reliability assessment of complex engineering systems and natural hazards, with the following main focus areas: (a) quantum machine learning, (b) quantum probabilistic inference, and (c) quantum optimization.

Read More

Deep Learning for Reliability and Risk

/
Deep Learning for Reliability and Risk

Modern engineered systems are generally configured with hardware, software, human, organizations, and their interactions. These features exponentially increase the complexity of systems and, in turn, boosting the possibility of system failure. Proper understanding and modeling of such system behaviors are important and urgent to ensure the reliability and safety of system design and operations. With the rapid advances on the Internet of Things (IoT), modern engineered systems tend to be increasingly instrumented with network-connected devices and massive quantities of multidimensional data have been generated from multi-sensor suites. These big machinery data have been recognized as a valuable resource, based on which one can leverage on deep learning techniques to uncover and explore hidden features to gain insights on a complex system performance. Deep learning has emerged as a powerful approach that enables handling big machinery data and sensor fusion for efficient risk and reliability predictive solutions under uncertainty. CRRE develops research and predictive solutions for the industry that seek to tackle the challenges such as (a) physics-based deep learning, (b) Bayesian deep neural networks, (c) generative models, (d) interpretability and trustworthy deep learning based prognostic solutions, (e) systems of systems fault diagnostics and prognostics.

Read More