The project will focus on the development of algorithms to make sense of complex sequences of data acquired in different points in space. The algorithms will be useful in a wide range of settings, but the project is particularly geared towards multiphase flow measurement – which is key to industrial applications where different kinds of substances are being transported inside a pipe, such as oil, water, and gas.
Understanding complex systems
Recent advances in measurement technology allow us to record time series from complex systems, including for example nervous systems in biological organisms, human economies, the behaviour of flocks of birds, and the climate.
Such complex systems are difficult to model and predict, but recording detailed temporal information about them and analysing the data with advanced algorithms can help us to understand their behaviour better.
Analysing such data, however, carries challenges. The fact that these systems are ruled by complex physical phenomena makes the data difficult to interpret using classical analyses.
To address these challenges, Dr Escudero will lead the development of algorithms to inspect data recorded over time from sensors scattered over different locations, to be modelled as networks.
Mathematically, a networks represents a collection of objects – in this case sensors – and links that show connections between them. An example is the railway network, where train stations (nodes) are connected by the tracks (links).
Dr Escudero’s team will apply their findings to a variety of scenarios, from measurement science to financial and climate data. The team will collaborate with Professor Chao Tan, a renowned expert in measurement science and flow phenomena at Tianjin University, China.
Commenting on the new research, Dr Escudero said “This data science project aims to produce a new generation of analysis techniques to reveal underlying nonlinear physical dependencies in complex data recorded over networks.”