This chapter covers the following topics:
To analyze process manufacturing and discrete manufacturing shop floor data, you must acquire data from various data sources such as ERP applications, Manufacturing Execution System (MES), Quality/Laboratory Information Management System (LIMS) as well as from shop floor sensor device enabled machines. The solution provides adaptors to collect data from many sources. Oracle partners and customers can extract structured data from external source systems and semi structured data from machines and equipment sensor devices and load them into the data lake in Oracle Cloud. Equipment and sensor device data is contextualized with equipment and work order information and summarized for analysis.
You can use out of the box integrations between Oracle Adaptive Intelligent Apps for Manufacturing and Oracle applications, such as Oracle E-Business Suite applications. These out of the box adapters are provided with Oracle Adaptive Intelligent Apps for Manufacturing and enable it to collect data from Oracle E-Business Suite applications. You can use REST services in Oracle Adaptive Intelligent Apps for Manufacturing to extract and load data from equipment, sensor devices, and other non-Oracle applications such as MES, LIMS or any custom data sources to Oracle Cloud. Data from these various Information Technology (IT) and Operational Technology (OT) systems can be uploaded for any historical period of time and for one or many products using periodic and incremental batch modes.
Structured data can be extracted from external source systems and loaded into the Oracle Adaptive Intelligent Apps for Manufacturing data lake in Oracle Cloud. The two types of data uploads that are supported are:
Case Record Data
Business Entity Data
Semi-structured data that can be acquired from shop floor sensor devices include:
Stream Data
Alert Data
Oracle Adaptive Intelligent Apps for Manufacturing provides adapters that enable integration with Oracle applications whose data is stored in an Oracle database. These adapters use Data Pump Import for the initial load and use Oracle GoldenGate Cloud Service for the incremental loads (see: Oracle Cloud Using Oracle GoldenGate Cloud Service) to replicate data from their source systems into the database used by Oracle Adaptive Intelligent Apps for Manufacturing. The following key entities are replicated from their source systems and used for Oracle Adaptive Intelligent Apps for Manufacturing model building and analysis:
Master data, such as items, categories, resources, resource instances, operations, routings, bills of materials, and recipes.
Reference data, such as work orders or batches, sales orders, purchase orders, and quality test specifications.
Transactional data, such as work order material and resource transactions, quality data, material move transactions, and work order completions.
Oracle Adaptive Intelligent Apps for Manufacturing uses comprehensive REST services to acquire data from non-Oracle applications. REST services extract data from other data sources, such as custom or third party Enterprise Resource Planning (ERP) applications, Manufacturing Execution Systems (MES), Quality and Laboratory Information Management System (LIMS) systems, and legacy systems, as well as from machines, equipment, and sensors. Oracle partners can help users to configure and map the data from other data sources into Oracle Adaptive Intelligent Apps for Manufacturing. The REST services load the extracted data into the data lake in the Oracle Cloud.
Case Record Data files capture historical work orders or assembly serial data. This upload of prepared data using the provided CSV templates can be used for quick analysis. The data is mined and prepared into flattened records and then imported into Oracle Adaptive Intelligent Apps for Manufacturing to obtain insights.
See: Importing Case Record Data, Oracle Adaptive Intelligent Apps for Manufacturing Data Ingestion User's Guide.
Business Entity Data files capture the key entities from external source systems. This detailed upload of business entities enables leveraging all AIMFG features including Insights, Predictions, Genealogy and Trace, and Factory Command Center. The Business Entity file upload captures data for the individual business entities like items, lots, departments, person, machines, receiving, work orders, quality, and so on and Oracle Adaptive Intelligent Apps for Manufacturing processes all the uploaded data together for analysis by understanding the underlying data structures.
Business Entity Data can be imported using the provided CSV templates, or REST services, or out of the box integrations like Oracle Data Pump or GoldenGate (for E-Business Suite only).
See: Importing Business Entity Data, Oracle Adaptive Intelligent Apps for Manufacturing Data Ingestion User's Guide.
Sensor Stream data (like temperature) and Alert data (like Machine Status such as up, down, idle, or in-use) from shop floor sensor devices can be imported into Oracle Adaptive Intelligent Apps for Manufacturing using CSV files. The sensor data is contextualized with the Business Entity data and summarized for analysis. Separate CSV templates are provided to capture sensor stream and alert data.
Before you can use a sensor device to acquire and process data, use the Create Sensor Device Mappings page to:
Enter the sensor device ID and attribute details such as manufacturer, model number, and serial number.
Map the defined sensor device to an ERP equipment and ERP equipment instance.
Map each sensor device stream parameter to an ERP equipment parameter and parameter Unit of Measure (UOM).
Set the stream sample interval for processing events.
Select a time series feature set name in production analysis usage for extracting features for model analysis.
Select a times series feature set name in machine event analysis usage for event identification process.
Specify the rules for equipment and work order contextualization.
Overview of Time Series Data:
Time Series data from the equipment sensor devices comprises of a sequence of values or events obtained over a period of time and recorded at specific intervals. You can define time series feature sets to extract features for model analysis or detecting events from time series sensor device stream data.
You can create time-series feature sets by selecting a specific time segment and a specific simple function, for example like standard deviation, or a Symbolic Aggregate approXimation advanced function. Symbolic Aggregate approXimation (SAX) is a symbolic representation for complex time series data.
See:
Oracle partners can help users to configure machine data acquisition systems, such as Supervisory Control and Data Acquisition (SCADA), Distributed Control Systems (DCS) and other gateway device systems, to extract machine sensor data in the CSV file format. You can then upload these sensor data files into Oracle Adaptive Intelligent Apps for Manufacturing in batch mode. The application processes the sensor stream data and alert data, contextualizes it with equipment and work order information, and summarizes the contextualized data for analysis.
See: Importing Sensor Devices Data, Oracle Adaptive Intelligent Apps for Manufacturing Data Ingestion User's Guide.
The following data ingestion methods are available to ingest data into Oracle Adaptive Intelligent Apps for Manufacturing:
Comma separated values (CSV) files can be used to import both structured and semi-structured data.
REST Services are provided to import data into Oracle Adaptive Intelligent Apps for Manufacturing using CSV files.
Oracle Data Pump and GoldenGate (for Oracle E-Business Suite only) are out of the box integration provided between E-Business Suite and Oracle Adaptive Intelligent Apps for Manufacturing enabling high-speed transfer of data and metadata from source database to target database.
Oracle Adaptive Intelligent Apps for Manufacturing provides specific CSV templates, and Data Ingestion user interfaces to upload Case Record Data, Business Entity Data, and Sensor Devices Data. From the Home page, click Insights or Predictions, then Data Ingestion. You can then select the specific user interface, depending on the data you would like to upload.
Note that you must define and be assigned to a Case Record Data organization to upload case record data files and to a Business Entity Data organization to upload the business entity data files. See Setting Up Data Ingestion, Adaptive Intelligent Apps for Manufacturing Data Ingestion User's Guide.
See the Oracle Adaptive Intelligent Apps for Manufacturing Data Ingestion User's Guide for detailed information on uploading and importing data from various manufacturing enterprise systems and machine sensor devices into Oracle Adaptive Intelligent Apps for Manufacturing.