ModelarData

At ModelarData, we apply state-of-the-art research to improve data transfer, advanced analytics, and machine learning for industrial IoT. We are currently developing ModelarDB, an enterprise-grade open-source time series management system that provides very efficient compression and fast analytics for vast data sets of time series from industrial IoT such as sensors in wind turbines and photovoltaic panels. Such equipment can generate multiple TBs of data which cannot be handled by traditional DBMSs. ModelarDB achieves huge compressions of time series by representing them by means of models instead of storing the raw data. The user can configure a tolerable error bound for the models, e.g., 1%, to allow better compression. The error bound can, however, also be set to 0 in which case the compression is lossless.