Analytics and advanced data processing

Each viewpoint will have unique analytic requirements. Engineers and front-line managers' analytic needs may be more functional in nature and may include real-time stream analysis, anomaly detection, and prescriptive analytics, that is, predicting when a part or machine needs maintenance. Middle and upper managers representing the business viewpoint may only need the aggregated data or calculated key performance indicators, as well as data from ERP, CRM, or other enterprise or departmental systems.

Advanced analytics involves the application of statistics and machine learning to analyze trends, make predictions, and provide dynamic feedback in real time. Advanced analytics can be categorized as follows:

  • Descriptive analytics: This analyzes what has happened and what is happening, and is typically used in historical reporting and operational dashboards
  • Predictive analytics: This applies statistical models and machine learning to predict outcomes and understand trends
  • Prescriptive analytics: This recommends or automates a course of action for achieving a goal or desired outcome through simulation

The goal of advanced analytics is to improve, or supplement, decision making or component operations by applying use case-appropriate analytic algorithms, such as clustering and regression analysis. Visualization of the analysis is a crucial factor for human decision making and can help provide a level of confidence in the resulting decisions.