Guidelines for Big Data analytics solutions, including Artificial Intelligence

This deliverable addresses the use case scenario and how AI is used in each of them More in detail it describes flows for automated visual inspection as well as for global asset monitoring including risk analysisThe specific aspects addressed by this deliverable areDealing with the different dimensions of the datadriven approach compared to the traditional approach especially regarding the dimension ofBig Data Leveraging from abundant data and continuous data inflow from sensorsModel of the physical assets Automatically updating mathematical models of the structures with sufficient data to go beyond the accuracy of physicsbased models Probabilistic models Combining AIMachine Learning and probability theoretical frameworksTools for dataset preparation especially to assist the model developer in selecting suitable data subsets through selection manipulation either visually or using program code of the dataset or through exploration validation of the selected subsets using a visual frontendDataset exploration and visualisation Setups to visually inspect the quality of the data suitable for modelling and dashboards for scenario analysisGeneric Machine Learning modelling Exploiting the potential of AI to verify existing models made during design find new patterns and detect anomalies from the technical norms