# PECOS **Domain:** Austin Research / Laboratories **Doc Type:** Research entity **Source basis:** Supplied research cluster with linked references; see the [[wiki/Austin Research Evidence Map|evidence map]] for verification scope. **PECOS**, the Center for Predictive Engineering and Computational Sciences at Oden, develops reliable computational predictions of complex physical systems. Its defining requirements are validated mathematical models, controlled numerical errors, and quantified uncertainty. It is the central laboratory-scale bridge between [[wiki/Predictive Science Academic Alliance Program|PSAAP]]’s mission and Oden’s wider mathematics of inference and control. The original 2008 PSAAP award placed [[wiki/Robert Moser|Robert Moser]], [[wiki/J. Tinsley Oden|J. Tinsley Oden]], and [[wiki/Omar Ghattas|Omar Ghattas]] within a predictive-science program grounded in [[wiki/National Nuclear Security Administration|NNSA]]’s simulation needs. The 2020 PSAAP III award extended the work toward exascale computation in collaboration with [[wiki/Sandia National Laboratories|Sandia]], [[wiki/Los Alamos National Laboratory|Los Alamos]], and [[wiki/Lawrence Livermore National Laboratory|Lawrence Livermore]]. ## Experiment and simulation The plasma-torch research couples fluid/plasma dynamics, electromagnetism, kinetics, radiation, numerical algorithms, and experiments. Measurements constrain the model; discrepancies expose numerical or physical-model limitations; revised experiments can reduce consequential uncertainty. [[wiki/Texas Advanced Computing Center|TACC]] provides the computing required to execute these calculations at scale. This **experiment → measurement → model → VVUQ → prediction → revised experiment** sequence supplies an especially concrete example of the executable loop. Reduced models address the additional requirement that a prediction arrive in time to affect a decision. Its transfer to another domain is a methodological opportunity whose accuracy must be established in that new domain. ## Connected entries [[wiki/Oden Institute|Oden Institute]] · [[wiki/Predictive Science Academic Alliance Program|Predictive Science Academic Alliance Program]] · [[wiki/Robert Moser|Robert Moser]] · [[wiki/Omar Ghattas|Omar Ghattas]] · [[wiki/J. Tinsley Oden|J. Tinsley Oden]] · [[wiki/Texas Advanced Computing Center|Texas Advanced Computing Center]] · [[wiki/Verification Validation and Uncertainty Quantification|Verification Validation and Uncertainty Quantification]] · [[wiki/Reduced-Order Modeling|Reduced-Order Modeling]] · [[wiki/Sandia National Laboratories|Sandia National Laboratories]] · [[wiki/Los Alamos National Laboratory|Los Alamos National Laboratory]] · [[wiki/Lawrence Livermore National Laboratory|Lawrence Livermore National Laboratory]] · [[wiki/Advanced Simulation and Computing|Advanced Simulation and Computing]] · [[wiki/Austin Executable Loop|Austin Executable Loop]] · [[wiki/J. J. Pickle Research Campus|J. J. Pickle Research Campus]] · [[wiki/National Nuclear Security Administration|National Nuclear Security Administration]] · [[wiki/Predictive Science|Predictive Science]] ## Sources and corpus context - [[research/The Austin Executable Loop|The Austin Executable Loop — master document]] - [UT Austin Wins $17 Million Grant from Department of Energy for Computational Research](https://news.utexas.edu/2008/03/07/the-university-of-texas-at-austin-wins-17-million-grant-from-department-of-energy-for-computational-research/) - [Predictive Science Research Gets Major Boost Thanks to the Department of Energy](https://news.utexas.edu/2020/10/05/predictive-science-research-gets-major-boost-thanks-to-the-department-of-energy/) - [PECOS — Center for Predictive Engineering and Computational Sciences](https://pecos.oden.utexas.edu/about.html) - [Predictive Engineering and Computational Sciences](https://oden.utexas.edu/research/centers-and-groups/predictive-engineering-and-computational-sciences/) **Cluster:** [[wiki/Austin Executable Loop|Austin Executable Loop]] · [[wiki/Austin Research Evidence Map|Austin Research Evidence Map]] <!-- BEGIN AUSTIN SURVEILLANCE FIELD --> ## Prediction confidence across the surveillance field The surveillance-field essay puts two requirements together: an estimate must be credible enough to use and fast enough to affect a decision. PECOS’s validation and uncertainty work addresses the first; reduced models address computational timing. [[wiki/Population State Estimation|Population estimates]] and [[wiki/Source Trust Tuple|source-trust assessment]] supply comparisons with different observation and error models. A transition to a new application requires validation at that application’s scale, not confidence inherited merely from the method’s defense pedigree. **Follow:** [[wiki/Austin Surveillance Field|Austin Surveillance Field]] · [[wiki/Population State Estimation|Population State Estimation]] · [[wiki/Reduced-Order Modeling|Reduced-Order Modeling]] **Source:** [[research/The Austin Surveillance Field|The Austin Surveillance Field]]. <!-- END AUSTIN SURVEILLANCE FIELD -->