High throughput experimentation enables researchers to obtain reliable and versatile data on chemical processes. These highly-dimensional datasets are input par excellence for advanced computer modeling or machine learning. 
Researchers from the KAUST Catalysis Center used the Flowrence high-throughput platform to obtain a dataset for the oxidative coupling of methane. This is an important reaction to produce basic chemicals from feedstock other than crude oil. The team worked out a robust principle component analysis methodology and kinetic modeling.
Their recent publication in the journal Chemical Engineering Science shows how they dealt with the high-throughput data curation and how their approach could be validated with the kinetic model for the oxidative coupling reaction of methane, which is regarded as a complex chemistry in this respect.