Overview
This chapter covers the following topics:
Overview of Oracle Adaptive Intelligent Apps for Manufacturing
Oracle Adaptive Intelligent Apps for Manufacturing collects, stores, and analyzes massive amounts of Operational Technology (OT) data coming from shopfloor systems such as equipment, machines, sensors, and test stations and then contextualizes it with Information Technology (IT) data coming from business applications such as Supply Chain Manufacturing (SCM), Enterprise Resource Planning (ERP), Human Capital Management (HCM), and Customer Relationship Management (CRM). Oracle Adaptive Intelligent Apps for Manufacturing then analyzes the data by applying machine learning, data mining, and artificial intelligence techniques to discover key patterns and correlations that affect manufacturing efficiencies and provides actionable predictive analytics to maximize yield, and minimize defects, scrap, cycle times, costs, etc. It also provides comprehensive capabilities for backward and forward tracing of products and processes within manufacturing and supply chain spanning manpower, machine, material, method and management aspects to facilitate rapid root cause, impact and containment analysis. The following graphic depicts this flow of data:
Oracle Adaptive Intelligent Apps for Manufacturing Data Flow
Oracle Adaptive Intelligent Apps for Manufacturing provides four modules with advanced analytical capabilities. These four modules, which you can access from the Home page, shown in the graphic below, are:
-
Insights
-
Predictions
-
Genealogy and Trace
-
Factory Command Center
Insights
Insights from historical data analysis enable business users to discover the hidden patterns between influencing factors and a production metric, known as the target measure. The insights are presented to the business user in an easy to use user interface. Business users can drill into a specific insight to gain comprehensive understanding about the influencing factors by visualizing the correlations and distributions from the historical data set. Insights are supported for process manufacturing , discrete manufacturing and serialized manufacturing use cases.
Use the Insights module to:
-
Create a dataset to prepare the context for analyzing historical data.
-
Create an Insights model for a target measure using selected features and algorithms.
-
Review and publish the insights.
-
Deploy the Insights model.
-
View the findings in aesthetic and self-explanatory visualizations.
Predictions
The Predictions module provides predictive alerts on a target measure in the current or later operations in the production cycle. The module alerts you to variations from the target early in the process based on predictive analysis.
Process Manufacturing users can access real time predictions of a product target measure for current and future operations.
Discrete Manufacturing users can access real time predictions of an assembly target measure for current and future operations.
Additionally, discrete serialized manufacturing users can view real time predictions of an assembly serial unit target measure for current and future operations.
Use the Predictions module to:
-
Create a dataset to prepare the context for analyzing historical data.
-
Create a Predictions model that helps to predict a target measure for a product using selected predictors and an algorithm.
-
Analyze the confusion matrix to evaluate model performance. The confusion matrix displays the accuracy of the model (actual vs. predicted target measure results) for various results classifications, such as high, on target, or low.
-
Deploy the Predictions model.
-
View predictions for the batches, work orders, or serial units currently in production.
Genealogy and Trace
The Genealogy and Trace module enables end-to-end tracking and analysis of material composition, test results, and processes undertaken at every operation step of the manufacturing process, which provides complete traceability of a partially or completely finished product. This traceability includes items procured (internally or externally), consumed, manufactured, outsourced, shipped, returned, repaired, serviced in the field, and re-shipped during a particular period.
The Genealogy and Trace module includes a Timeline Viewer and a Network Viewer to enable traceability using a simple and intuitive visualization of the entire product genealogy. Use these viewers to drill down through objects such as equipment, sales orders, purchase orders, user defined entities, and other details to trace the impact of an object in the supply chain network. You can easily switch back and forth between the Network and Timeline Viewers as needed from any object in the view.
Use the Timeline Viewer to view:
-
All events associated with a work order, an equipment instance, a lot number, or a serial number occurring over time.
-
Only certain steps or types of events associated with a work order, equipment instance, lot number, or serial number.
-
The details of any associated event.
Use the Network Viewer to view:
-
A purchase order, lot, serial number, work order, or sales order and the associated entities.
-
Details about the relationship between two objects, such as the lot quantity received for a purchase order.
-
A serialized item or lot controlled item and then trace the item from the supplier to its consumption in a work order and its shipment to a customer.
Factory Command Center
The Factory Command Center displays different metrics and prediction alerts related to the overall factory/organization and classified under the 5 Ms of the factory: Manpower, Machine, Material, Management, and Method. Examples of metrics and alerts you can expect to see are:
-
Manpower - information related to the operators assigned to the running work orders, such as operators not clocked in, not reported, and skills mismatch.
-
Machine - the machines that are down, scheduled for maintenance, and idle.
-
Material - material with a shortage or expiring.
-
Management - work order-related delays, such as work order start or finish delays, operation delays, or work orders on hold.
-
Method - changes to the planned method of operation, such as changes in operation or activity duration, unplanned operations, or material, processing, and resource exceptions.
The above metrics and alerts are reported for current work orders with a status of Unreleased or Pending, Released, In Progress, and On Hold.