Credibility has been defined as a sufficient degree of belief in the validity of a model to justify its use for research and decision making (Rykiel, 1996). In other words, as the satisfaction of the end-user in terms of accuracy, robustness, and uncertainty of the simulation. In the context of data-intensive science and technology (industry 4.0, Internet of Things, ) industrialists need credible simulation procedures embedded into decision making protocols. Data assimilation (the process by which experimental observations of the system are incorporated into the model) appears therefore as a key ingredient in this process. The purpose of this project is to develop novel computational techniques to simulate industrial and medical problems with two critical aspects for the user. On one hand, it is essential to dynamically incorporate observations of the system into the decision-making pipeline. On the other hand, it is crucial for the end-user to incorporate state-of-the art simulations with quantifiable credibility indicators into the daily decision-making routine. The ultimate goal is to put forward the basis for a new generation of numerical tools for the connected industry and autonomous devices such as cars. To achieve this, CrediblE is focused on three specific industrial challenges (IPs): radar sensor modelling for autonomous cars, augmented reality for laparoscopic surgery assistance and manufacturing in the connected industry. These three problems, although apparently unconnected, share many methodological similarities. They require fast and reliable modern techniques based on model order reduction, manifold (machine) learning or artificial intelligence, just to name a few. Moreover, all need incorporating observations (data assimilation). Finally, the credible simulation for a decision-making process needs accuracy assessment, robustness & sensitivity, as well as uncertainty quantification (UQ). Note that UQ also includes other common ingredients of the problems at hand: the need for fast (real-time) solution of inverse problems arising from the experimental observations, parameter estimation and data assimilation. Therefore, CrediblE aims at providing tools to face the Social Challenge on Economy and Digital Society (#7 from the list of those identified by the Spanish RTD Strategy). Incidentally, making progress in the three IPs (autonomous cars, surgery and manufacturing) is having impact in three other Social Challenges: Health, Demographical Change and Welfare (#1); Clean, Safe and Sustainable Energy (#3) and Integrated and Sustainable Smart Transport (#4).
Garikapati, H.; V. Verhoosel, .; van Brummelen, .; Zlotnik, S.; Diez, P. Computational geosciences Vol. 23, num. 1, p. 81-105 DOI: 10.1007/s10596-018-9784-y Date of publication: 2019-02 Journal article
Peral, M.; Kiraly, A.; Zlotnik, S.; Funiciello, F.; Fernandez, M.; faccenna, C.; Vergés, J. Tectonics Vol. 37, num. 9, p. 3285-3302 DOI: 10.1029/2017TC004896 Date of publication: 2018-09 Journal article