Oracle Data Platform for Manufacturing

Manufacturing plant data consolidation

 

Optimize efficiency and lower risk with consolidated, real-time data

Today’s manufacturers must understand how efficiently all their lines are running across multiple plants—they need to know immediately when a problem occurs, not five or ten minutes after the fact. However, this is also one of their biggest challenges because their ability to do this relies on real-time access to data from multiple remote locations that may have limited or sporadic internet connectivity. To solve this problem, we need to push machine learning (ML) and data acquisition to the network edge.

Simplify decision-making at the edge

We can configure Oracle Data Platform to solve this challenge by including Oracle Roving Edge Devices (REDs). Each RED is designed to capture, store, run, manage, and gain insight from data, giving manufacturers the ability to automate the decision-making process and management of manufacturing equipment at the edge. Oracle Data Platform for manufacturing also includes anomaly detection capabilities, which can be used to address manufacturing line disruptions and provide maintenance-related insights to improve mitigation and remediation.

The following architecture demonstrates how Oracle Data Platform supports plant data consolidation by deploying advanced analytics and machine learning at the edge to identify anomalies, perform smart data collection, and provide real-time operational information.

plant data consolidation diagram, description below

This image shows how Oracle Data Platform for manufacturing can be used to consolidate plant data. The platform includes the following five pillars:

  1. 1. Data Sources, Discovery
  2. 2. Ingest, Transform
  3. 3. Persist, Curate, Create
  4. 4. Analyze, Learn, Predict
  5. 5. Measure, Act

The Data Sources, Discovery pillar includes two categories of data.

  1. 1. Business records data comprises warehouse management and inventory optimization data, ERP (Oracle E-Business Suite, Fusion SaaS, NetSuite) data, and MES planning and scheduling data.
  2. 2. Technical input data includes sensor, camera, and device (IoT) data and data from PLM, SCADA, and manufacturing applications.

The Ingest, Transform pillar comprises four capabilities.

  1. 1. Batch ingestion uses Oracle Data Integrator and OCI Data Integration.
  2. 2. Streaming ingest uses Kafka Connect.
  3. 3. Custom integration uses Oracle WebLogic Server on VMs.
  4. 4. RED sync transfer uses a local daemon.

Batch ingestion connects unidirectionally to the serving data store.

Streaming ingest and custom integration connect unidirectionally to the outbound transfer area.

Additionally, RED sync transfer unidirectionally connects to the inbound transfer area.

The Persist, Curate, Create pillar comprises four capabilities.

  1. 1. The serving data store uses MySQL and Oracle DB server.
  2. 2. Batch processing/Spark processing uses OCI GoldenGate Stream Analytics.
  3. 3. The outbound transfer area uses OCI Object Storage.
  4. 4. The inbound transfer area uses OCI Object Storage.

These capabilities are connected within the pillar. Batch/Spark processing is unidirectionally connected to the serving data store.

The outbound transfer area is unidirectionally connected to batch/Spark processing.

Three capabilities connect into the Analyze, Learn, Predict pillar:

The serving data store connects unidirectionally to the analytics and visualization capability and bidirectionally to the anomaly detection capability. The outbound transfer area connects unidirectionally to the anomaly detection and RED sync transfer capabilities.

The inbound transfer area connects unidirectionally to the anomaly detection capability.

The Analyze, Learn, Predict pillar comprises three capabilities.

  1. 1. Analytics and visualization uses Oracle Analytics Server.
  2. 2. Anomaly detection uses a model trained centrally and deployed locally as PMML.
  3. 3. RED sync transfer uses a local daemon.

The anomaly detection capability is unidirectionally connected to the analytics and visualization capability within the pillar.

Three capabilities are connected to the Measure, Act pillar. The analytics and visualization capability is unidirectionally connected to local dashboards and reports and also local predictions. The anomaly detection capability is unidirectionally connected to local predictions, and the RED sync transfer capability is unidirectionally connected to an additional use case.

The Measure, Act pillar captures how the consolidated plant data can be used. These potential uses are divided into four groups.

  1. The first group includes local dashboards and reports.
  2. The second group includes local predictions.
  3. The third group includes applications.
  4. The fourth group contains an additional use case, which is operational efficiency and performance.

The three central pillars—Ingest, Transform; Persist, Curate, Create; and Analyze, Learn, Predict—are supported by Oracle Roving Edge Device(s).



There are four main ways to inject data into an architecture to enable manufacturers to easily understand operational efficiency and performance.

  • A custom integration from Oracle Integration Repository lets us integrate data—both structured and unstructured—from various sources, allowing for interactions with devices, custom APIs, and so on. The data can be ingested from any application development type (for example, standalone Java or Python code, Oracle WebLogic Server–based applications, or Kubernetes-based applications). Data will be stored in object storage for further refinement, for outbound transfer, or to feed AI models.
  • The RED data sync is an efficient and simple way to transfer ML models from a central location (for example, your object storage repository of trained models in Oracle Cloud Infrastructure (OCI)) to the edge. In this use case, the edge definition would have the RED colocated with other machinery within the plant itself. New versions of models are stored in “standalone” Predictive Model Markup Language (PMML) format. The local daemon will perform an update when a new model is discovered and automatically push it to the RED. The RED data sync is also a great way to transfer all the data collected by different REDs throughout the day (for example, relevant anomalies, signals, and so on) to your central location, most likely to object storage on OCI. This data will then be used for operational reporting and ML model training. The volume of data involved in these RED data sync processes will determine your requirements for edge-to-data center telco or satellite bandwidth.
  • Batch ingestion uses Oracle Data Integrator, a comprehensive data integration solution that covers all data integration requirements from high-volume, high performance batch loads to event-driven, trickle-feed integration processes and SOA-enabled data services. While real-time needs are evolving, the most common extract from ERP, planning, warehouse management, and transportation management systems is a batch ingestion using an extract, transform, and load or extract, load, and transform process. These extracts could be frequent, as often as every 10 or 15 minutes, but they are still bulk in nature as transactions are extracted and processed in groups rather than individually. OCI offers different services to handle batch ingestion; these include the native OCI Data Integration service or Oracle Data Integrator running on an OCI Compute instance. Depending on the volumes and data types, data could be loaded into object storage or loaded directly into a structured relational database for persistent storage.
  • Analyzing data in real-time from multiple sources can help provide manufacturing companies with valuable insights into their operational efficiency and overall performance. Oracle Data Platform uses streaming ingestion to ingest data streams from several ISA-95 Level 2 systems, such as supervisory control and data acquisition (SCADA) systems, programmable logic controls, and batch automation systems. Streaming data (events) will be ingested and some basic transformations/aggregations will occur before the data is stored in object storage. Streaming analytics can be used to identify correlating events, and identified patterns can be fed back (manually) for a data science examination of the raw data. While traditional analytics tools extract information from data at rest, streaming analytics assesses the value of data in motion, i.e., in real time.

Data persistence and processing is built on three components.

  • In the serving data store, data will be managed by Oracle Database Server or MySQL for data processing. The serving data store provides a persistent relational tier often used to serve data directly to end users via SQL-based tools. It also functions as the serving layer for specialized analytics.
  • All data retrieved from data sources in its raw form (as a native file or extract) is captured and loaded into object storage to be used in current or future ML model training. Cloud object storage is the most common data persistence layer for our data platform, and it serves as both the inbound transfer area and the outbound transfer area. It can be used for both structured and unstructured data.
  • With object storage as the primary data persistence tier, OCI GoldenGate Stream Analytics is the primary processing engine. Batch processing involves several activities, including basic noise treatment, missing data management, and filtering based on defined outbound datasets. Results are written back to various layers of object storage or to a persistent relational repository based on the processing needed and the data types used.

The ability to analyze, learn, and predict is built on two technologies.

  • Analytics and visualization services deliver descriptive analytics (describes current trends with histograms and charts), predictive analytics (predicts future events, identifies trends, and determines the probabilities of uncertain outcomes), and prescriptive analytics (proposes suitable actions, leading to optimal decision-making). Oracle Analytics Server provides the functionality to deliver descriptive analytics related to operational reporting and prescriptive analytics. Additionally, ML models can be embedded directly into the Oracle Analytics Server data flow. Oracle Analytics Server is designed to run on-premises and provides dashboards, reporting, alerting, self-service data preparation, and end user–driven machine learning algorithms. Oracle Data Platform for manufacturing is completely open and flexible, so, if desired, you could use third-party tools for this instead.
  • Alongside the use of advanced analytics, ML models are developed, trained, and deployed to support anomaly detection. OCI Anomaly Detection is an AI service that makes it easier for developers to build business-specific anomaly detection models that flag critical incidents, speeding up detection and resolution. These models will be trained at the central location and deployed in PMML format to be executed locally as Java or Python code.

Automate decision-making to increase profitability

Oracle Data Platform lets manufacturers get the greatest value from all their available data while simplifying and streamlining data access and storage. The ability to push data collection and ML scoring to the edge through Oracle Roving Edge Devices helps manufacturers make better business decisions that are informed by accurate data that’s always available when they need it, allowing them to increase efficiency and production while lowering costs.

Related resources

開始使用 Oracle Modern Data Platform

試用 20 多項永遠免費雲端服務,並可在 30 天内試用其他服務

Oracle 所提供的 Free Tier 包含無時間限制使用 20 多項服務 (如 Autonomous Database、Arm Compute 和 Storage) 以及 300 美元的免費額度,可試用其他雲端服務。取得詳細資料,並立即註冊您的免費帳戶。

  • Oracle Cloud Free Tier 包含哪些項目?

    • 2 個 Autonomous Database,每個 20 GB
    • AMD 和 Arm 運算 VM
    • 200 GB 總區塊儲存體
    • 10 GB 物件儲存體
    • 每月 10 TB 輸出資料傳輸
    • 10 多項永遠免費服務
    • 300 美元的免費點數,可在 30 天内試用其他服務

深入瞭解逐步指引

透過教學課程和實作實驗室來體驗廣泛的 OCI 服務。無論您是開發人員、管理員還是分析人員,我們都可以協助您瞭解 OCI 的運作方式。許多實驗室是在 Oracle Cloud 免費層或 Oracle 提供的免費實驗室環境上執行。

  • 開始使用 OCI 核心服務

    此研討會中的實驗室涵蓋 Oracle Cloud Infrastructure (OCI) 核心服務介紹,包括虛擬雲端網路 (VCN) 以及運算和儲存服務。

    立即開始 OCI 核心服務實驗
  • Autonomous Database 快速入門

    在此研討會中,您將會瞭解使用 Oracle Autonomous Database 的步驟。

    立即開始 Autonomous Database 快速入門實驗
  • 以試算表建立應用程式

    本實驗將引導您將試算表上傳到 Oracle Database 表中,然後根據此新表格建立應用模組。

    立即開始此實驗
  • 在 OCI 上部署 HA 應用程式

    在本實驗中,您將會在 Oracle Cloud Infrastructure (OCI) 中的兩個運算執行處理上部署 Web 伺服器,並使用負載平衡器以高可用性模式 (High Availability) 進行設定。

    立即開始 HA 應用程式實驗

探索 150 多種最佳實務設計

瞭解架構師與其他客戶如何部署各種工作負載,從企業應用程式至 HPC,以及從微服務到資料湖。透過「Built & Deployed」影片系列,瞭解最佳實務,聽聽其他客戶架構師的分享,您也可使用我們的「按一下即可部署」功能來部署許多工作負載,或者從我們的 GitHub 存放區自行部署。

熱門架構

  • Apache Tomcat 搭配 MySQL Database Service
  • Kubernetes 上的 Oracle Weblogic,搭配 Jenkins
  • 機械學習 (ML) 與人工智慧環境
  • Arm 上的 Tomcat,搭配 Oracle Autonomous Database
  • 日誌分析與 ELK 堆疊
  • HPC 與 OpenFOAM

瞭解 OCI 可以助您節省多少成本

Oracle Cloud 的定價簡單明瞭,在全球各地保持一致的實惠價格,而且支援廣泛的使用案例。若要預估您的費率,請查看費用預估工具,並依照您的需要設定服務。

體驗與衆不同之處:

  • 1/4 的外送頻寬成本
  • 3 倍運算性價比
  • 每個區域的定價同樣實惠
  • 定價實惠,且無須長期履行合約

聯絡業務代表

有興趣深入瞭解 Oracle Cloud Infrastructure 的資訊嗎?讓我們的專家助您一臂之力。

  • 我們的業務代表可以爲您解答以下問題:

    • 哪些工作負載在 OCI 上的執行效果最佳?
    • 如何從 Oracle 的整體投資中獲得最大收益?
    • OCI 如何與其他雲端運算提供者比較?
    • OCI 如何支援您的 IaaSPaaS 目標?