Smarter analysis of industrial data

Important results from the project

Target was to identify new relations between process and product, develop a more efficient process analytics methodology and to find a new business model within automated development. New relations were found but deeper analysis is needed. A more efficient process analytics methodology adapted for the industrial challenge was developed and valuable conclusions regarding possibilities for more efficient development work and prerequisites for new business models within automated development were drawn.

Expected long term effects

The software for connecting time dependent machine data to quality data developed within the project has identified several very interesting relations and will be used for further in-depth analysis of these. Also the process model optimization work yielded interesting results to be further developed. Furthermore, the conclusions regarding more efficient analysis of industrial data will be communicated and discussed within the companies prior to future strategies regarding data analytics.

Approach and implementation

Data extraction and analytics were done in an iterative manner since data was more difficult than expected to extract and analyze. This was because the IT-infrastructure, data storage and software were not adapted for massive data analysis but also because the amount of quality features was small compared to the amount of process data. This fact gave limitations regarding the data analytics methodology and also increased the demands on data analytics competence and domain knowledge of the analyzed process. Industry should increase analytics competence and efforts on data storage.