Operational Data

4 Key excerpts on "Operational Data"

• 

  eBook - ePub

  Smart Manufacturing

  Concepts and Methods

  • Masoud Soroush, Michael Baldea, Thomas F. Edgar(Authors)
  • 2020(Publication Date)
  • Elsevier

    (Publisher)

  The right side of the pyramid represents activities typically performed by support centers, which utilize data to improve operations and enable business initiatives (e.g., improved asset management) using noncontrol software applications. These initiatives provide actionable information to all functions of the plant for productivity improvement. The operational metrics can be calculated by using proper data classification and aggregation, at a high level of detail. This strategy enables faster communication and collaboration within the functional teams at the plant and within the global business via the cloud. At this level the business and time contexts are added to the process measurement. The data and events are analyzed to determine the best course of action for the plant to realize tangible benefits.

  5: Using operations data for enterprise or division-level data analytics

  The integration of operational analysis and BI is the next level. At this level, plans are constantly being adapted by defining the best targets to optimize all process areas. This activity is typically performed by a division-wide or corporate team. The data management and automation are what allow these strategies to be successfully implemented.

  The collection and contextualization of real-time data are essential to provide input to more sophisticated offline predictive analytics and data analysis tools. Aggregating asset and time context to the data and operational state of the data allows significant production events to be processed automatically, storing needed information at the desired level of detail. Fig. 3 shows the process of refining the data to enable the transformation of raw sensor data so that it can be used for enhanced analytics, predictions, and advanced nonproduction visualization analysis. This nonproduction analysis typically uses IT visualization tools that are less time-series oriented.

  Fig. 3 Process of cleansing and classifying production data for enhanced business analytics. (Courtesy O.A. Bascur, OSIsoft LLC.)

  Today, operations and production data may be required for business analysis. The word information is derived from the Latin forma. Information requires the data to have a structure-based and additional context. In additional to the physical context of the data, it requires the time dependence related to the operational states of the process unit and all the process attributes. Fig. 3

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Business Intelligence Guidebook

  From Data Integration to Analytics

  • Rick Sherman(Author)
  • 2014(Publication Date)
  • Morgan Kaufmann

    (Publisher)

  • Operational systems live in the here and now, whereas data warehousing must support the past, present, and future. Operational systems record the business event as is, whereas data warehousing tracks changes in dimensions—products, customers, businesses, geopolitical, account structures, and organizational hierarchies—so that information can be examined as is, as was, and as will be.

  • Operational Data typically contains a relatively short time span, whereas analytical data is historical. A business needs to perform period-over-period analysis or examine trending using historical data.

  • The data for many of the attributes that the business wants to analyze is neither needed nor available in an operational system.

  • Operational Data is spread out over many source systems, making it hard to bring together and analyze. The more sources you have, the more data integration you will need. • Every enterprise, no matter how large or small, must perform data integration to ensure that its data is consistent, clean, and correct.

  • There are many business algorithms used to transform data to information outside of operations systems. Finance, sales, marketing, and other business groups each must transform the data into the business context they need to perform their work.

  • There are both enterprise-wide and business group-specific performance measures of key performance indicators (KPIs) that need to be derived outside of operational systems.

  There have been numerous times when vendors proclaim that data warehousing is no longer needed. Over the years, we have heard them talking about middleware, virtual data warehouses, conformed data marts, enterprise information integration (EII), enterprise application integration (EAI), service oriented architectures (SOA), data virtualization, and real-time access from every generation of BI tools. In fact, it is a recurring theme. There is no “silver bullet” that helps an enterprise avoid the hard work of data integration. Information that is clean, comprehensive, consistent, conformed, and current is not a happenstance; it requires thought and work.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Oracle Business Intelligence and Essbase Solutions Guide

  • Rosendo Abellera, Lakshman Bulusu(Authors)
  • 2016(Publication Date)
  • Auerbach Publications

    (Publisher)

  Take for instance the example where the transactional system would take an order and in doing so would be able to gauge a corresponding result as to the supply level of materials needed to create that product that was ordered. So in other words, with this capability, items in inventory can be ordered in advance to manufacture the product. The inventory level of raw materials used and consumed for manufacturing is properly gauged and measured. Through this process operational business intelligence is able to determine that raw materials used in the manufacturing of the product and determine whether materials are running low in supply. So as the order is taken, the system simultaneously reorders more raw material to satisfy any future orders. This is just but one example of how Operational BI could get the right data to the right person at the right time and consequently gain an advantage in keeping the business of manufacturing running smooth by decreasing the chance of running out of material and not lose an opportunity to sell more products due to a shortage of raw material for manufacturing. It may seem very simple but at this time of writing it has taken decades to effectively be able to source data from transactional systems and move and transform it in a way that it is usable for analysis within the course of a day. This time-sensitive capability of business intelligence or analytics of an operational system that can offer a real competitive advantage to any organization is Operational BI.

  With this goal in mind, we now turn our focus in this chapter to being able to deliver that system or solution that can provide the necessary operational business intelligence and analytics to achieve the goal previously described. We now reveal the practitioner’s secrets necessary to succeed in delivering a holistic and complete solution for Operational BI.

  Operational BI for near real-time data analytics and analysis can be centralized around an Operational Data Store (ODS) complementing the EDW that does function as a persistent store of nightly or inter-day data. The ODS can deliver minute-by-minute data from operational systems by way of creation of semantic views based on a logical data model and the data for the views being materialized on the fly. These views eliminate the need to move data physically in addition to giving the flexibility to transform and reload data on the fly. Another approach can be in-memory caching of a table of metrics/KPIs for dashboards in OBIEE. Operational BI can benefit from both of these approaches and deliver near-time data for analytics and analysis.

  7.5  Delivering Operational BI

  The underlying data structures and data architecture supporting the BI solution are the enablers most often overlooked when developing or implementing a solution entailing Operational BI. For structured data in a corporate environment, proper data modeling is vital in developing an Enterprise Data Warehouse. This is the “secret sauce” that at times, whether because of lack of expertise or even knowledge, is sometimes dismissed and is the cause of much of the failure that plagues the industry. In the past and even with some of today’s new tools and technologies, vendors have incorporated strategies and approaches attempting to avoid having to model the data. Although it is perhaps permissible for unstructured data to be in a setting for discovery, not modeling the data in a structured environment and solution such as for an enterprise data warehouse is not good practice. It would be analogous to building a house without a blueprint. Today’s best practices for building BI solutions have reached the maturity to fully understand the limitations and capabilities of various different data structures and architectures as they relate to the various different categories and types of BI solution, ranging from Operational BI to Analytical BI and beyond.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Data Warehousing Fundamentals for IT Professionals

  • Paulraj Ponniah(Author)
  • 2011(Publication Date)
  • Wiley

    (Publisher)

  This category of data comes from the various operational systems of the enterprise. These normally include financial systems, manufacturing systems, systems along the supply chain, and customer relationship management systems. Based on the information requirements in the data warehouse, you choose segments of data from the different operational systems. While dealing with this data, you come across many variations in the data formats. You also notice that the data resides on different hardware platforms. Further, the data is supported by different database systems and operating systems. This is data from many vertical applications.

  In operational systems, information queries are narrow. You query an operational system for information about specific instances of business objects. You may want just the name and address of a single customer. Or, you may need the orders placed by a single customer in a single week. Or, you may just need to look at a single invoice and the items billed on that single invoice. In operational systems, you do not have broad queries. You do not query the operational system in unexpected ways. The queries are all predictable. Again, you do not expect a particular query to run across different operational systems. What does all of this mean? Simply this: there is no conformance of data among the various operational systems of an enterprise. A term like an account may have different meanings in different systems.

  The significant and disturbing characteristic of production data is disparity. Your great challenge is to standardize and transform the disparate data from the various production systems, convert the data, and integrate the pieces into useful data for storage in the data warehouse. It is really the integration of these various sources that provide the value to the data in the data warehouse.

  Internal Data

  In every organization, users keep their “private” spreadsheets, documents, customer profiles, and sometimes even departmental databases. This is the internal data, parts of which could be useful in a data warehouse.

  Sign up to read

  Learn more about book