data governance · 2019-08-26 · data governance council data owner data stewards (business and...

37
© Fraunhofer ISST DATA GOVERNANCE Prof. Dr.-Ing. Boris Otto 28 September 2018 Dortmund public Bildquelle: guinnessworldrecords.com (2017). · 1

Upload: others

Post on 25-Jun-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

DATA GOVERNANCE

Prof. Dr.-Ing. Boris Otto 28 September 2018 Dortmund

public

Bildquelle: guinnessworldrecords.com (2017).

· 1

Page 2: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

CONTENT

A Brief History of Data Governance

Data Governance in Business Ecosystems

The IDS Approach to Data Governance

public· 2

Page 3: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Around the millennium change Data Governance increasingly received attention as a response to compliance risks

Image sources: infrapark-baselland.com (2018), bruecken.deutschebahn.com (2018). Logos from company websites and Wikipedia (2018).

public

Financial Regulations

Bankruptcy of energy giant Enron due to fictional financial reporting

In the course of this process, Arthur Andersen found guilty of obstruction of justice for shredding thousands of documents

The company surrendered its CPA license on August 31, 2002, and 85,000 employees lost their jobs

Governmental Regulations

»Leistungs- und Finanzierungsvereinbarung(LuFV)« links funding of Deutsche Bahn to quality of infrastructure inventory

Direct relationship between quality of data and financial situation

Environmental Regulations

Chemical spill into the river Rhine in 1986 at Sandoz plant in Basel-Schweizerhalle

No data about nature and implications of chemical substances spilled

· 3

Page 4: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Business drivers for Data Governance were – and still are – multifold and affect the company as a whole

public

Group Level

Division 2Division 1 Division 3

Business units

Business processes

Locations

Business units

Business processes

Locations

Business units

Business processes

Locations

Compliance to regulations

360 degree view of the customer

Integrated and automated business processes

»Single Source of the Truth« for business reporting

Smooth business integrations

· 4

Page 5: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data quality evolves over time according to a »jigsaw« pattern

Legend: Data quality issues.

Data Quality

TimeProject 1 Project 2 Project 3

public· 5

Page 6: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Reasons for poor data quality are manifold – as the example of Bayer CropScience shows

NB: For background on the case study see Ebner et al. (2011).

public

Data Quality Issues

Employees Data Maintenance

DQ Management Standards Organization

Training and education inadequate

Data quality not integrated in performance management systems

Various software solutions in place

Master data can be edited in target systems

No integrated software support

Data maintenance not harmonized on global level

No data qualitymetrics

No continuous data quality monitoring

No binding rules, standards, operating

procedures

Too many local rules, exceptions

No“Data Governance”

Missing business responsibilities

· 6

Page 7: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Corporate life is hard without Data Governance

Image source: Strassmann (1995).

public· 7

Page 8: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data Governance and Data Quality Management are closely interrelated

Source: Otto (2011).

public

Legend: Goal Function Data.

Data

Governance

Data Quality

Management

Maximize

Data Quality

Maximize

Data Value

Data Resource

Data Resource

Management

is sub-goal of

supports supports

is led by is sub-function

of

are object of is object of

are object of

· 8

Page 9: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

A strategic resource is a source of competitive advantage

Strategic Resource

V Value

R Rarity

I Inimitability

N/ONon-substitutability

Organization

Source: Barney (1991); Makadok (2001).

public

VRIN/VRIO Framework

Resources

»all assets, capabilities, organizational processes, firm attributes, information, knowledge, etc. controlled by a firm that enable the firm to conceive of and implement strategies that improve its efficiency and effectiveness«

Capabilities

»special type of resource, specifically an organizationally embedded non-transferable firm-specific resource whose purpose is to improve the productivity of the other resources possessed by the firm«

Resource-Based View of the Firm

· 9

Page 10: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Despite its intangible nature, industrial data has a value which can be quantified

Source: Moody & Walsh (1999).

public

Number of users

Share of value

100% Data

Tangible Goods

Tangible Goods

ValueData

Usage Time

Potential value

Data

Data quality

Value

100%

Data

Integration

Value

Data

Volume

Value

Data

· 10

Page 11: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Many examples exist demonstrating the applicability of valuation procedures in the data domain

Source: Otto (2012); Otto (2015), Zechmann (2017).

Company Industry Country Data domainValuationapproach

Value per record

Retail USCustomer data including shopping profile

Market value 1.6 EUR

Social Network US User data Market value 225 USD

Automation and drives

DE Master data on partsProduction costs

500 to 5.000 EUR

Agrochemical CH Material master dataUse/income value

184 CHF

public· 11

Page 12: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data Governance aims at allocating decision rights for the management and use of data within an organization

Source: Otto (2011).

Data Governance Organization

Data Governance Goals Data Governance Structure

Formal Goals

Business Goals

Ensure compliance Enable decision-making Improve customer satisfaction Increase operational efficiency Support business integration

IS/IT-related Goals

Increase data quality Support IS integration (e.g. migrations)

Functional Goals

Create data strategy and policies Establish data quality controlling Establish data stewardship Implement data standards and metadata management Establish data life-cycle management Establish data architecture management

Locus of Control

Functional Positioning

Business department IS/IT department

Executive management Middle management

Hierarchical Positioning

Organizational Form

Centralized Decentralized/local Project organization Virtual organization Shared service

Roles and Committees

Sponsor Data governance council Data owner Data stewards (business and technical)

public· 12

Page 13: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data Governance is typically established as an enterprise-wide virtual organization – as the example of BOSCH shows

Source: Bosch (2008).

public

Master Data

Owner n

Executive Management

Master Data ManagementSteering Committee

Group Division/Central Function

Accountability onBusiness Unit Level(Data Maintenance)

IT Projects

IT Platforms, IT Target Systems

Overall Accountability(organizational level) Master Data

Owner A

Master DataDomain 1

Master DataDomain n

Report

Governance

Working GroupTeam of Experts

ConceptsConcepts

Governance

… …

e.g. Vendor Master Data Chart of Accounts

Inte

rdiscip

lina

rilysta

ffed

Master Data Officer

Master Data Officer

· 13

Page 14: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

A data quality index is an effective performance management tool at Bayer CropScience

Source: Ebner & Brauer (2011).

84

86

88

90

92

94

96

98

100

11/2009 01/2010 03/2010 05/2010 07/2010 09/2010 11/2010 01/2011

Material Master Data Quality Index

Asia Pacific

Europe

Latin America

North America

[%]

public· 14

Page 15: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Johnson & Johnson has reached a six sigma data quality level

Source: Otto (2013).

99,503

94,586

95,50696,102

95,77896,312

95,656

89,855

91,629

96,324 96,383

97,433

95,417

99,135

99,885 99,971 99,993 99,999

84

86

88

90

92

94

96

98

100

02.15.11 04.15.11 06.15.11 08.15.11 10.15.11 12.15.11 02.15.12 04.15.12 06.15.12

Data Quality Index

Data Quality Index

public· 15

Page 16: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Five key principles lead to excellence in master data governance

Source: Otto & Österle (2015).

Capture Data at the Source

Enter Data »First Time Right«

Measure to Manage

Build up a Data Governance Capability

Scale Capabilities Globally

public· 16

Page 17: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Life’s good with Data Governance

Image source: Strassmann (1995).

public· 17

Page 18: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Developed by the Competence Center Corporate Data Quality, the Data Excellence Model (DXM) defines building blocks for data management

Source: Competence Center Corporate Data Quality (2017).

public

GOALS ENABLERS RES ULTS

D A T A

S T R A T E G Y

P E O P L E , R O L E S &

R E S P O N S I B I L I T I E S

P R O C E S S E S &

M E T H O D S

D A T A

L I F E C Y C L E

D A T A

A P P L I C A T I O N S

D A T A

A R C H I T E C T U R E

P E R F O R M A N C E

M A N A G E M E N T

B U S I N E S S

C A P A B I L I T I E S

D A T A

M A N A G E M E N T

C A P A B I L I T I E S

B U S I N E S S

V A L U E

D A T A

E X C E L L E N C E

· 18

Page 19: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Smart Data Engineering is model-based, method-oriented approach for building up an effective Data Resource Management capability

Defining the data strategy

Assigning roles and responsibilities for core data domains

Managing data as an economic good

Designing a consistent data architecture for the digitalized enterprise

Controlling the business benefit contribution of the data resource

public· 19

Page 20: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

CONTENT

A Brief History of Data Governance

Data Governance in Business Ecosystems

The IDS Approach to Data Governance

public· 20

Page 21: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data has become a strategic enterprise resource

Legend: MRP – Manufacturing Resource Planning; ERP – Enterprise Resource Planning.

public

Data as a Process Result Data as a Process Enabler Data as a Product Enabler Data as a Product

Information systems have been used since the 1960s and 1970s to support enterprise functions, but data wasn‘t shared between functions, let alone enterprises.

With the proliferation of MRP and ERP systems in the 1980s and 1990s data enabled end-to-end business processes such as order-to-cash, procure-to-pay, make-to-stock etc.

Since the millennium change, data has increasingly become an enabler of innovative product-service-systems and integrated solutions.

Recently, data marketplaces emerged offering data APIs at a volume or frequency based fee.Data has become a product in its own right.

Mainframe Computing Enterprise Systems Electronic Business Data Economy

· 21

Page 22: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

In the era of digitalization, companies must develop their Data Management from »Defense« to »Offense«

Source: DalleMulle & Davenport (2017).

public

Defense Offense

Key ObjectivesEnsure data security, privacy, integrity, quality, regulatory compliance, and governance

Improve competitive position andprofitability

Core ActivitiesOptimize data extraction, standardization,storage, and access

Optimize data analytics, modeling,visualization, transformation, andenrichment

Data Management Orientation

Control Flexibility

Enabling Architecture Single Source of Truth Multiple Versions of the Truth

· 22

Page 23: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data Intelligence Hub

Data sharing platform

Data sovereignty and security

The data economy is here

Sources: Deutsche Telekom (2018); HERE (2018); CDQ (2018).

public

HERE Tracking Cloud

Community approach to data management

Using the power of many

Deutsche Telekom HERE Corporate Data League

· 23

Page 24: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Sharing data is a prerequisite for ecosystems

Image sources: Johns Hopkins University (2016), Umweltbundesamt (2016), Smellgard, Schneider & Farkas (2016), urbanmanagement.nl (2017).

Data Sharing

Energy

Health Care

Material Sciences

Manufacturing and Logistics

»Smart Cities«

Sharing of material information along the entire product life cycle

Shared use of process data for predictive asset maintenance

Exchange of master and event data along the entire supply chain

Anonymized, shared data pool for better drug development

Shared use of data for end-to-end consumer services

public· 24

Page 25: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data sovereignty is a prerequisite for innovative business models in variousdomains

Image sources: perm4.com (2017); hccs.edu (2017); dvz.de (2017).

Health Care Patient Data

Use purpose

Anonymization

System constraints

Personalized medicine

Better healthcareservices

Domain Data Usage Conditions Value Potential

ProductionProduct Data

Process Data

Usage frequency

Usage types

Use purpose

Innovative productionnetworks

»Production as a Service«

Automotive Planning and Risk Data

Use purpose

Expiration date

System constraints

Better risk management

Less production bottlenecks

public· 25

Page 26: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

The role of Data Governance differs between Offense and Defense Data Management…

Image source: ebay (2018).

public

Defense Offense

Scope Enterprise-internal Ecosystem, Customer

Ownership Setting data standards Executing property rights

Stewardship Quality Curation

Organization Hierarchy Market, Community

Data Flows Internal between application systems Data value chains in networks

Usage Access Rights Usage Rights

Economics Cost and Use Value Market value

· 26

Page 27: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

CONTENT

A Brief History of Data Governance

Data Governance in Business Ecosystems

The IDS Approach to Data Governance

public· 27

Page 28: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

The IDS Reference Architecture Model responds to the most important issues in data sharing

Source: PwC (2017). The International Data Spaces (IDS) Association publishes the IDS Reference Architecture Model (IDS-RAM). The Industrial Data Space is a vertical application of the IDS-RAM.

57%worry about revealing valuable data and business secrets.

59%fear the loss of control over their data.

55%feel inconsistent processes and systems as a (very) big obstacle.

32%fear that platforms do not reach the critical mass, so that data exchange will be interesting.

InteroperabilityData SovereigntyTrust and Security Join us!

Today

IDS Approach

public· 28

Page 29: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data sovereignty is needed for effective Supply Chain Risk Management

OEM»Tier 1« Supplier

Risk Management

Supplier Management

• Contact person

• Risk type

• Risk location

• Affected parts

• Affected sub-suppliers

• Capacities and inventory levels

• Contact person

• Parts demand

• Inventory levels

Use contextRisk management

ConditionDeletion after 3 days

Use contextSupplier management

ConditionDeletion after 14 days

public· 29

Page 30: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data sovereignty is needed for innovation in the pharmaceutical industry

Pharma Company

Usage context

Clinical research

Anonymization

Data record must

consists of at least

150 individual

anonymized data

sets

University Hospital

Patient Management

Smart Drug Development

• Health data

• Medication plan

• Electronic case records

public· 30

Page 31: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data sovereignty is a prerequisite for flexible and dynamic production networks

“Production as a Service” Provider

OEM

ProductionPlanning and

Control

• CAD data

• Configuration parameters

• Production volume

• Usage time

• Temperature data

• Certificates

Usage contextMaintenance, no forwarding

ConditionOperator anonymous

Maintenance

Usage contextMachine type

ConditionDelete CAD data after first use

public· 31

Page 32: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Usage conditions for data are multifold

Dimension Specification Example

Geo-information

Coordinates 51.493773, 7.407025, radius 1km

Geo polygon

ZIP code 44227

Country code DE

Expiration date Absolute date December 24, 2017

Anonymization

Role, function

Usage purposePositive list Use for machine configuration

Negative list Not for marketing use

PropagationAllow, deny

Allow on a fee Yes, with 20 percent surplus charge

Number of uses Absolute figure Once

Deletion

System constraints

public· 32

Page 33: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

The Industrial Data Space provides an architecture for the sovereign exchange of data

Legend: IDS Connector; Usage Constraints; Non-IDS Communication.

public

Industrial Data Cloud

IoT Cloud

Enterprise Cloud

Data Marketplace

Company 1 Company 2 Company n + 2Company n + 1Company n

Open Data Source

IDS

IDS IDS

IDS

IDS IDS

IDS

IDSIDS

IDS

IDSIDS

IDS

IDS

IDS

IDS

IDS

· 33

Page 34: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

The Industrial Data Space forms an ecosystem around the sovereign exchange of data

Quelle: IDS Reference Architecture Model Version 2.0 (2018).

public· 34

Page 35: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Data Governance activities are distributed to the different roles in the IDS ecosystem

NB: Activities in brackets are to be discussed.

public

IDS Role Data Governance Activity IDS Software Component

Data Owner/Provider

Define usage constraints for data resources Publish metadata (incl. usage constraints) to broker Transfer data with usage constraints linked to data Receive information about data transaction from Clearing House Bill data (if required) (Monitor policy enforcement)

IDS Connector

DataConsumer/User

Use data in compliance with use constraints IDS Connector

Broker Match data demand and supply Broker Software

Clearing House Monitor and log data transactions and data value chains (Monitor policy enforcement) (Perform data accounting)

Clearing House Software

App StoreProvider

Offer data governance and data quality services App Store Software

· 35

Page 36: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

Prof. Dr.-Ing. Boris Otto

Fraunhofer ISST · Executive DirectorTU Dortmund · Faculty of Mechanical Engineering

[email protected] · [email protected]

https://de.linkedin.com/pub/boris-otto/1/1b5/570

https://twitter.com/drborisotto

https://www.xing.com/profile/Boris_Otto

http://www.researchgate.net/profile/Boris_Otto

http://de.slideshare.net/borisotto

Please get in touch!

public· 36

Page 37: Data Governance · 2019-08-26 · Data governance council Data owner Data stewards (business and technical) · 12 public ... Optimize data analytics, modeling, visualization, transformation,

© Fraunhofer ISST

DATA GOVERNANCE

Prof. Dr.-Ing. Boris Otto 28 September 2018 Dortmund

public

Bildquelle: guinnessworldrecords.com (2017).

· 37