Managing Big Multidimensional Data - For Energy And Beyond

Torben Bach Pedersen
TrackOpening Plenary -- Main Conference Auditorium
DescriptionMultidimensional database concepts such as cubes, dimensions with hierarchies, and measures have been a cornerstone of analytical business intelligence tools for decades. However, the standard data models and system implementations (OLAP) for multidimensional databases cannot handle “Big Multidimensional Data”, very large amounts of complex and highly dynamic multidimensional data that occur in a number of emerging domains such as energy, transport, logistics, as well as science. This talk will focus on how to manage Big Multidimensional Data, including modeling, algorithmic, implementation, as well as practical issues.
The main application domain considered in the talk is energy data management. Energy data come in many forms, e.g., time series of energy consumption and production data, data about available energy flexibilities, price data, weather data, etc. Typically, most of these things are forecasted both at the long, medium, and short term, before they are finally measured and captured, a complex procedure which adds further challenges to the management. Especially, integrating large amounts of renewable energy like wind and solar into the energy production is challenging, since it requires to shift energy demand to where and when the supply is, rather than the other way round, an example of the intelligent energy management infrastructure known as the smart grid. The talk will present data management challenges and solutions from the MIRABEL and Totalflex projects, which develop a unique "data-driven" approach to the smart grid, allowing to predict, capture, aggregate, and utilize the inherent flexibilities in energy demand and supply, with a focus on flexible demand and renewable energy supply.

Presentation documents

All talks