Uncertainty Quantification in Reliability Prediction of Aging Systems

Professor SankaranMahadevanhas more than twenty-five years of research and teaching experience in reliability and risk analysis methods, design optimization, structural health monitoring, and model verification, validation and uncertainty quantification (V&V and UQ) methods. His research has been extensively funded by NSF, NASA, FAA, DOE, DOD, DOT, NIST, General Motors, Chrysler, Union Pacific, American Railroad Association, and Sandia, Idaho, Los Alamos and Oak Ridge National Laboratories. His research contributions are documented in more than 180 journal articles and numerous other publications, including two textbooks on reliability methods. He has directed 40 Ph.D. dissertations and 23 M. S. theses, and has taught many industry short courses on reliability and risk analysis methods. His awards include the NASA Next Generation Design Tools award (NASA), the SAE Distinguished Probabilistic Methods Educator Award, and best paper awards in MORS Journal and the SDM and IMAC conferences.

Professor Mahadevanobtained his B.S. from Indian Institute of Technology, Kanpur, M.S. from Rensselaer Polytechnic Institute, Troy, NY, and Ph.D. from Georgia Institute of Technology, Atlanta, GA.

This talk will discuss current research directions and opportunities regarding uncertainty quantification in long-term reliability assessment and decision-making in engineering systems. Model-based simulation is attractive for systems that are too large and complex for full-scale testing. However, model-based simulation involves many approximations and assumptions, and thus confidence in the simulation result is an important consideration in risk-informed decision-making. Sources of uncertainty are both aleatoryand epistemic, stemming from natural variability, information uncertainty, and modeling approximations. The presentation will draw on illustrative problems in aerospace, mechanical and civil engineering to discuss (1) the effects of data uncertainty and model uncertainty (both due to model form assumptions and solution approximations) on long-term reliability assessment of multi-physics, multi-scale systems; (2) dynamic Bayesian networks for continuously integrating heterogeneous information from multiple sources (models, tests, experts) in multiple formats throughout the life-cycle of the system, including damage sensing and repair activities; and (3) opportunities for uncertainty-informed decision-making throughout the life cycle of engineered systems, such as accelerated testing, material design, and risk management.