The volume, variety and velocity of data and associated uncertainties differ widely in various life cycle phases. The data should also be readily accessible for safety and risk assessments. However, the connectivity, availability and cybersecurity of data transmission, in terms of on-location data storage versus enterprise cloud, remains one of the key challenges for all organizations.
The data and digitization cross-functional team will address the data analytics and digitalization needs for three application areas throughout their entire operational life cycles and those required by other core areas. Data analytics research will produce the deliverables for predictive maintenance, performance optimization, system risk identification and mitigation, human reliability and environmental safety assessment.
Team Leads
Data and digitization cross-functional team leads are:
- Ali Mosleh (University of California – Los Angeles)
- Steven DiMarco (Texas A&M University)
- Myank Tyagi (Louisiana State University)
- Kelly Rose (National Energy Technology Laboratory)
Focus Areas
- Data standards and models: Novel data models, standards and protocols for both structured and unstructured data will be undertaken to build robust and accurate machine learning models for different objectives.
- Data collection, visualization and archival: Internet-connected sensors and mobile and edge computing will be used to collect real-time synoptic data on wind, waves, currents and seafloor movements for any installation, maintenance planning, operations and response operation. Cloud technologies will be used for data repository archival and retrieval for all offshore engineering systems through proper organizational approvals, advanced methods and technologies for data summarization and visualization will be developed.
- Data analytics and uncertainty quantification: Methods and tools for assessing predictive model parameters with sparse data will address storage needs of different types of datasets from different life phases of offshore engineered systems requiring specific data-driven models. Other teams, such as Cross-Functional Team 1 and Cross-Functional Team 3, will leverage such models to fulfill their tasks.
- Data mining: Descriptions of past incidents and near-misses will be leveraged through automated text mining. Advanced data mining techniques for various datasets and artificial intelligence will be used collaboratively with human experts for further analysis.
