Data analysis and optimization

The data analysis field is concerned with statistical analysis of large data sets and predictions based on those statistics. Industrial Applications of data analysis include root cause analysis, deviation detection, and sales forecasting. Optimization, on the other hand, is the science of finding optimal (or acceptable) solutions within a given set of constraints. Typical applications of optimization theory include production planning and logistics.

SICS does research in both areas, with a special focus on combining them for industrial applications. There are many situations when you want to re-optimize a system based on statistical analysis of some parameters. One current example is state-based maintenance, where service maintenance of some machinery is scheduled based on the actual state of the machine, as opposed to e.g. number of hours since last service interval. In this case, data analysis techniques are used to determine the state of different machines, and consequently which machines need to be serviced, while optimization techniques are used to continuously reschedule e.g. productions plans for the entire factory, taking service plans into consideration.

A specific area of optimization techniques in which SICS has a long history is logic programming, and in particular the programming language Prolog. SICStus Prolog was developed at SICS and is the world-leading Prolog platform for demanding applications.

More information
Industrial Applications and Methods group
SICStus Prolog