PGE Enterprise Information Management EIM Strategy Sendil Thangavelu

PG&E Enterprise Information Management (EIM) Strategy Sendil Thangavelu Lead Principle, Information Architecture PG&E

Executive Summary Current Situation • Changing Business Landscape: Changing regulations, Customer expectations & Smart Grid requirements around data will need new capabilities and systems with reduced latencies and time to market • Multiple BI systems: Multiple BI systems with data fragmented, light self service foot print, resulting in inconsistent Information, high TCO and low adoption • Integration: Point to Point Integrations. Some existing Integration systems are outdated and in some cases out of support. • Lack of Best Practices: Metadata, Master Data, Data Governance and holistic Information Architecture that help drive consistency, reduce delivery cycles and cost are non existent • Information centric Initiatives: Although related, often initiatives are implemented in silo adding to cost and fragmentation. Implementing data centric initiatives are proving to be expensive • Organizational paradigms: Functional group specific business processes and requirements ignore other cross functional enterprise impacts • Conclusion: There is significant need to rationalize data management systems and introduce innovative capabilities. 2

Executive Summary Recommendations • Develop holistic Information Management capability: Implement linked, incremental, foundational information capability that can be leveraged by LOB initiatives. Create additional capabilities in a phased, business driven manner. Some of the capabilities have a steep learning curve for both IT and business. Start sooner rather than later. • Establish Information Management as a priority: With the help of business stakeholders establish that Information is an Enterprise Asset. Form a Business-IT Steering committee with Senior Management to prioritize initiatives and track progress • Consolidate BI capabilities: Identify overlapping and redundant systems and consolidate. Define standards for future BI Platform • Move Beyond silo Reporting to Intelligent Enterprise: Implement cradle to grave, pro active Data Management • Establish Data Governance: Implement Pragmatic, nimble, purpose driven, repeatable framework and capability • Track Accountability: Reduce applications that are developed for short term goals. Track accountability for lifetime system maintenance cost. Enhance the existing governance structures for overall DSM. Next Steps • Get executive buy in to move forward with Information Management Foundational Phase • Prioritize Initiatives in Partnership with the Business • Define a roadmap for a “managed evolution” rather than big bang approach to adding functionality. • Show tangible results to the business 3

Scope of EIM capabilities Information strategies from conceptual value to operational impact 4

‘As-is’ Capabilities and Gap Information strategies from conceptual value to operational impact 5

As –is Key Information Challenges Issues Fragmented and Outdated Integration • Many interfaces (often redundant) cause higher costs Data distributed along fragmented application landscape • Inconsistent data • Cross functional data usage complex • Manual reconciliation processes • Multiple Data Warehouses Risks § Implementation of new functionality and app interfacing more complex and expensive § Difficulty in implementing crossfunctional projects § Complexity of various point-to-point interfaces almost not manageable § Cost and effort increases while issues continue to persist § Multitude of interfaces drive maintenance cost • Future Cross-Functional or data intensive business requirements would be hard to manage or implement • Complexity of managing data will increase significantly with data intense projects • With further development of data complexity, consistency becomes unmanageable 6

As-is Key Information Challenges Issues Multiple Technology stacks • Not necessarily different capabilities No footprint of Foundational capabilities • Master Data Management • Enterprise Data Quality and Governance • Meta Data Management No Enterprise Information Architecture New capabilities need to be stood up over time • Advanced Visualization • Complex Event Processing • BI as a Platform with consistent and complementing capabilities Risks • Increased TCO Licensing costs Resources Training Evolution and upgrades § Increased project costs due to Siloed and repeated efforts § Inconsistencies across projects § Fragmented & Redundant efforts § Multiple independent projects to address similar ‘Information needs’ § No IA governance on these projects leading to increased costs and further fragmentation § Smart Grid will create a data deluge § Acquiring, managing and converting data into actionable, reliable Information will need these capabilities to be rolled out in phases 7

Key Information Challenges Issues Organizational Issues • Project versus Enterprise mind set • Information intensive Projects implemented largely based on outsourced advice. • Shelf ware of Software products • Skill set gap/readiness to deploy new capabilities • Due to lack of in house skills sourcing and support model needs to be evaluated Risks § Continue to propagate redundant projects and assets § Information is inherently cross functional as such § Outsourced advise is a function of skill sets, not the best solution 8

Future State Information Requirements Organic Data Volume Growth § Ability to handle a significant increases in the number of operational data sources and associated data volume § Increasing reliance on data analytics and visualization capabilities due to significant increases in data volume § Devices with processors and two way communication that will enable collection of more information, decision making and coordination. § Higher, two way collaboration and business process integration between users, businesses, individual customers and a variety of technology systems, resources and intelligent devices. 9

Future State Information Requirements Business Requirements § Increasing need to move, secure, analyze and act on Information for a wider range of stakeholders and significantly reduced latency § Deployment of data for use by an increasing number of stakeholders § Increasing range of data latency and availability requirements § Utilization of operational data to make real-time decisions as well as for planning, scheduling and dispatch § Greater integration of operational and business system data Regulatory Requirements § Improved data security and an increase in user authorization levels / schemes § Increasing need to provide data for 3 rd party reporting (e. g. , regulatory reporting) § Compliance with CPUC mandated Open ADE § Compliance with FERC 2004 Access controls to reports around usage data 10

Future State Information Requirements Technology Requirements • Information Systems must be architected and designed to be adaptive and resilient to autonomous, independent, potentially unexpected or non-responsive behavior of the new participants. Example Distributed Generation 11

Target Set of capabilities Based on future State Requirements and the identified Information challenges, a Target set of capabilities need to be stood up in a phased manner. Linked, Incremental capability build out with tangible Business benefits are being proposed. While some efforts seem large, they can be implemented in a smaller scale, yet with an Enterprise view to iron out issues after which they can be propagated. Once established, these capabilities can: • • • Be leveraged by Line of Business initiatives These initiatives will also assure consistent, reliable Information Be re used in multiple projects 12

EIM Capabilities-Phased approach Information Architecture Phase III • • • DW/BI Rationalization BI Unified Platform Complex Event Processing Analytics Advanced Visualization Train of thought analysis Information Architecture–Phase II • Enterprise Data Integration with Mash ups- Information as Service Paradigm • Multi Domain Master Data Management (Incubator of many EIM disciplines • Enterprise Data Layer • SOA and Enterprise Service Bus Capability Phase III Capability-Phase II Information Architecture-Foundational Phase I • • Enterprise Semantic Model Enterprise Meta Data Management Enterprise Data Profiling and Quality Enterprise Data Governance Industry Standards (CIM) Information Lifecycle Management Best Practices Foundation- Phase I 13

EIM Capabilities-Phased approach Information Architecture-Foundational Phase I • • Enterprise Semantic Model Enterprise Meta Data Management Enterprise Data Profiling and Quality Enterprise Data Governance Industry Standards (CIM) Information Lifecycle Management Best Practices Information Architecture Phase III Information Architecture–Phase II • Enterprise Data Integration with Mash ups- Information as Service Paradigm • Multi Domain Master Data Management (Incubator of many EIM disciplines • Enterprise Data Layer • SOA and Enterprise Service Bus • • • DW/BI Rationalization BI Unified Platform Complex Event Processing Analytics Advanced Visualization Train of thought analysis 14

Benefits of EIM Key Best Practice Capabilities Direct Benefits Stakeholder Benefits • Increased data accuracy, completeness, Master Data, Meta Data and Data Quality conformity, consistency, and integrity • Standard for data retrieval • Effective remediation processes • Future-proof people, process and tech to meet Data Integration Architecture Complex Event Processing Capabilities uncertain regulatory/industry factors • Data as a Service • Data-Store once use many times • Higher predictability and reliability • Align system processes with business processes • Reduce upstream workload volumes • Ability to transform large amounts of data to Advanced Visualization & Analytics Capabilities useful, comprehensible information • Improve customer relationships through targeted demand response programs • Enhance environmental and regulatory compliance through more effective tracking • Achieve greater network reliability and resilience through real-time performance updates Customer Increased reliability, increased quality of available information and reduced cost to serve Regulators Streamlined information gathering process and reporting Shareholders Reduction in overall cost / improvement in EPS 15

EIM Capabilities-Foundation phase Information Architecture Phase III • • • DW/BI Rationalization BI as a Platform Complex Event Processing Analytics Advanced Visualization Train of thought analysis Information Architecture–Phase II • Enterprise Data Integration with Mash ups- Information as Service Paradigm • Multi Domain Master Data Management (Incubator of many EIM disciplines • Enterprise Data Layer • SOA and Enterprise Service Bus Capability Phase III Capability-Phase II Information Architecture. Foundational Phase I • Enterprise Semantic Model • Enterprise Meta Data Management • Enterprise Data Profiling and Quality • Enterprise Data Governance • Industry Standards (CIM) • Information Lifecycle Management • Best Practices Foundation Phase I 16

Foundational Phase • Enterprise Semantic Model It is a model driven approach to managing Data, Information, Intelligence and Integration. It helps us understand how different pieces of information relate to each other in a consistent manner. It helps us achieve consistency from a conceptual model level all the way to run time artifacts 17

Enterprise Semantic Integration Benefits of Semantic Integration Enable broad data interface integration §Forces “semantic coherency” across all interoperable data interfaces Easily “Plug-in” additional data interfaces §Describe new interfaces in terms of a “business-like” conceptual model Lingua-franca for the business §Business, Analysts, developers, architects, data stewards can understand §Data Governance, Business Processes, Risk Visibility enabled Easily “Plug-in” additional data interfaces §Describe new interfaces in terms of a “business-like” conceptual model 18

Semantic Modeling Process 19

Enterprise Semantic Model Eco System 20

Foundational Phase Enterprise Meta Data Management Metadata is data about data. Describing a resource with metadata allows it to be understood by both humans and machines in ways that promote interoperability and re use. Metadata is structured information that describes, locates and makes it easier to retrieve, use, or manage an information resource. Types of Meta data Include: Business, Technical & Operational §What does this data mean ? §Where did it come from ? §How did it get there ? §Why is my report showing different data than your report? Who’s data is right ? 21

Foundational Phase Benefits • • • Increased confidence in data Assert Lineage, Quality and Fit for purpose Foster Discovery, Self service, adoption, sharing and Re use Helps machine to machine interaction such as Data Integration Reduce support needs and costs 22

Metadata Management Environment 23

Metadata at PG&E - Example Data Element: Productive Time Business Name: Productive Time Data Definition: Employee wages paid while the employee is at work … Abbreviation: Prd. Tm Data Source: SAP/Time Keeping Business Rule: Productive time (physical time at PG&E) can be non-billable for emails, meeting … Mouse over or right click to see Metadata “pop-up” with information about a specific data element Mouse over or right click to see metadata “pop-up” 24

Foundational Phase Enterprise Data Quality Data quality is an assessment of fitness of the data to serve its purpose in a given context. Aspects of Data Quality include Accuracy Completeness Timeliness Relevance Reliability Viability of business decisions are contingent on good data. . . Good data is contingent on an effective approach to Data Quality Management 25

Foundational Phase Enterprise Data Quality approaches Reactive: addresses problems that already exist deal with inherent data problems, integration issues, merger and acquisition challenges Proactive: diminishes the potential for new problems to arise Governance, roles and responsibilities, quality expectations, supporting business practices, specialized tools. Both approaches are needed. Profiling and quality management should be taken as upstream as possible in the data creation process 26

Data Quality-Iterative implementation approach 27

Data Profiling Architecture 28

Enterprise Data Governance Is an Organizational capability that oversees the use and usability of Data. involves people, process and Technology It Benefits • Increase consistency & confidence in decision making § Decrease the risk of regulatory fines § Improve data security § Achieve consistent information quality across the organization § Designate accountability for information quality § Semantic modeling will lend itself to Data Governance 29

Data Governance Paradigm Shift 30

Standards and Information Management The Smart Grid ecosystem will require a wide variety of information to be exchanged, managed, accessed analyzed. Standards specify object models that are the basis for efficient exchanges of Information between applications within and among grid domains. Broad implementation of these standards will enhance interoperability of applications and reduce the time and expense required to integrate new technologies and systems. Standards are a moving Target for Information Management. Certifications process is still nascent. At the core of many IEC standards is the IEC Common Information Model (CIM). § CIM has been officially adopted to allow application software to exchange information about the configuration and status of an electrical network § Some of the standards such as IEC 61850(Substatation Automation), IEC 61968 (Distribution) and IEC 61970 (Transmission), 60870 (Exchange of Information between control centers) are series with multiple parts, where some parts may be appropriate, or may only be in a proposed or draft form § Domain models provided by the CIM may be leveraged by PG&E as starter inputs for Enterprise Semantic Model 31

Logical Relationship amongst Standards 32

Data Integration and Possible Standards 33

Information Lifecycle Management The policies, processes , practices, services and tools used to align the business value of Information with cost efficient and appropriate Infrastructure from the time information is created to its final disposition Source: SNIA • Information has value, and that value changes over time • Older DOES NOT necessarily mean lower value for Information • A key Objective of ILM is to ensure cost of ownership to be commensurate with value of Information 34

EIM Capabilities-Phased approach Information Architecture Phase III • • • DW/BI Rationalization BI as a Platform Complex Event Processing Analytics Advanced Visualization Train of thought analysis Capability Phase III Information Architecture–Phase II • Enterprise Data Integration with Mash ups- Information as Service Paradigm • Multi Domain Master Data Management (Incubator of many EIM disciplines • Enterprise Data Layer • SOA and Enterprise Service Bus Capability-Phase II Information Architecture-Foundational Phase I • • • Enterprise Semantic Model Enterprise Meta Data Management Enterprise Data Profiling and Quality Enterprise Data Governance Industry Standards (CIM) Information Lifecycle Management Foundation Phase I 35