Large injection molding machine to make good use confidential data, we need to break through the core technology categories

The first is how the injection molding machine data management problems. After the massive injection molding machine machine type data such as time series, spatial and temporal high-speed data acquisition is complete, it needs to survive, this relates to data valid packing, compression, issue placed. Data survive it is to be used, which requires support queries to quickly locate the data application requirements, which in turn is an example of how to establish efficient spatiotemporal data indexing problems.

Better manage data stored, the next question is how to support a variety of analysis. Did people know the actual analysis, the analysis is much more than a bunch of algorithm development issues. Algorithm is only a small part of the work, most of the work is based on the understanding of the business problem to select the data needed to understand the characteristics of the data, and then design an appropriate model and algorithm based on the characteristics. This understanding of the characteristics of intermediate data big machine data is very difficult. Because the machine data can not be an intuitive understanding of the human need to interact with feature projects. In addition, from the perspective of models and algorithms, machine data is often a physical world system of perception results, and the physical world there are many mechanistic principles exist, such as mechanical fields of mechanics, metallurgy fields of chemistry, so the injection molding machine analysis of large-scale machine data requires an organic combination of mechanistic models and statistical models. Another often overlooked issue is data quality problem - how to grasp the quality of the data, how to fix data quality.

Angle talk application, how to more easily access and use of data analysis, especially for experts in the field. In the multi-source heterogeneous data, mask data integration problems associated with such experts in the field do not need to understand the complexity of big data technology and programming.