Composite Dev. Research at SAP - Stanford...

19
Composite Dev. Research at SAP Anne Hardy SAP Research, Palo Alto

Transcript of Composite Dev. Research at SAP - Stanford...

Page 1: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

Composite Dev.Research at SAPAnne Hardy

SAP Research, Palo Alto

Page 2: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 2

In current research projects, we cooperate with >200partners

50 Universities30 research organizations80 technology partners30 industrial users10 public sector users

Innovations emerge from ecosystems, not from singleplayers

We want to better leverage North AmericanUniversities.. Such as Stanford, …

SAP Research – Network

Page 3: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 3

Practical Research Methodology

Research

Research Pilot

Research Demo

Product Transfer

Product Enhancement

Criteria• Potential market• Customer input• Innovation• Impact to SAP

Criteria• Potential market• Customer input• Innovation• Impact to SAP

Page 4: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 4

Real World Awareness, Miniaturization, and the Internet ofThings

Recent advances in miniaturization, sensor & communicationtechnology, and new materials drive for a new computing paradigmand an explosion of information

Page 5: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 5

The Information Explosion

0

200

400

600

800

1000

1200

2005 2006 2007 2008 2009 2010

Ex

ab

yte

s

Six-fold Growth

Exabytes Created, Captured, Copied in a Year

Source: IDC, “The Expanding Digital Universe, March 2007

• Chevron – 2 Terabytes per day• London Traffic Surveillance – 64 Trillion Bits/day

Page 6: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 6

The Information Explosion

~ 65% Consumer Generated~ 85% Incurring EnterpriseResponsibility~ 90% Unstructured Data

Key Questions• How do we make sense of this data?• Is there unlocked business value in this data?

Page 7: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 7

Green Technology Momentum

Key Questions• Where can customers and SAP provide value?• What are the new business models?• What new processes and technologies will be required?

Page 8: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 8

The Rise of the Mobile Individual Business User (MIBU)

Worldwide Converged Mobile Device Market: Who Buys the Phone

Corporate purchasing driving use of individual-owned mobiledevices for Mobile Enterprise Applications

Key Questions• What mobile platforms and technologies

are required to support mobile Enterprise apps?

Page 9: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 9

The New Generation of Worker

Comfortable with the “Web 2.0”world

For communication, collaboration,information gathering…

Blurred line between work and“lifestyle”Low tolerance for laggingbusiness IT

Key Questions• What are their expectations of the SAP’s of the world?• How will they define business processes?

Page 10: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 10

SAP Research: Trend-driven Innovation

The NewGenerationof Worker

InformationExplosion

Rise ofthe MIBUGreen

Tech

Real World Awareness/Internet of Things

The SAP Research Visionand Research Areas

Page 11: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 11

SAP Research 2007: Vision and Research Areas

Technology/Platforms Applications

Service EcosystemsIndustrialization of SoftwareDevelopment

Service Delivery PlatformInternet of Things & Real-WorldAwarenessEnd-to-End SimplicityAdvanced Web TechnologiesSeamless Cross-OrganizationalInteroperabilityOpen Source & SecurityActive DatabasesMobility PlatformVirtualization/ParallelismComposite DevelopmentSemantics and Ontologies

Future ManufacturingWeb 2.0 ApplicationsService IndustryPublic SectorEnergy Distribution, Mgmt. AndConservationComposite ApplicationsProduct Lifecycle ManagementService Applications

Business AspectsGlobal Research Areas

Universities, Strategic Industrial Partners, Lighthouse Customers, ConsortiaGlobal Research Ecosystem

Vision: “Towards Business in Networked Economies”• Rapidly changing business designs and business models• Next generation of Web-related services and technologies• Architectures of participation and collective intelligence• IT-structures removing technology barriers to evolutionary

business process transformation

Page 12: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 12

SAP Research North Americas Research Fields

Risk Mitigation: Open Source

Mobility Platform

Palo Alto

Montreal

Internet of Things and RealWorld Awareness

Advanced Web Technologiesand Web 2.0 Applications

Continuous InformationSensemaking

Shanghai

Virtualization/Parallelism

Model driven Composite Dev.

Page 13: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 13

Composite Dev. - Motivation

The globalization trend and dynamics of today's business operations imposemany new requirements on software application development.

One of them is how to composite or extend an application by leveragingexisting software components and enabling the quick development ofnew/missing functions.

SAP FM

Non-SAP service

New function

Page 14: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 14

Typical Composite Dev. cycle

End User(LOB)

1.Define 2. Configure &Orchestrate

3. Develop 4.Deploy

BusinessOwner(LOB)

Developer(IT)

Administrator(IT)

Implement new services or UIcomponents and bring theminto modeling environment as“black box”

BusinessAnalyst (IT)

Map application landscape to designatedIT landscape also complete technicalconfiguration (property settings)

Business Configureof BP/App usingmodeling interface

Identify gaps tocomplete application(i.e. services needed,missing UI elements)

Prototype BusinessProcess, UI for eachstep and role as wellas Informationmodeling).

Select services forprocess and UI wherepossible

ValidateEnd User(LOB)

1.Define 2. Configure &Orchestrate

3. Develop 4.Deploy

BusinessOwner(LOB)

Developer(IT)

Administrator(IT)

Implement new services or UIcomponents and bring theminto modeling environment as“black box”

Implement new services or UIcomponents and bring theminto modeling environment as“black box”

BusinessAnalyst (IT)

Map application landscape to designatedIT landscape also complete technicalconfiguration (property settings)

Map application landscape to designatedIT landscape also complete technicalconfiguration (property settings)

Business Configureof BP/App usingmodeling interface

Business Configureof BP/App usingmodeling interface

Identify gaps tocomplete application(i.e. services needed,missing UI elements)

Identify gaps tocomplete application(i.e. services needed,missing UI elements)

Prototype BusinessProcess, UI for eachstep and role as wellas Informationmodeling).

Select services forprocess and UI wherepossible

Prototype BusinessProcess, UI for eachstep and role as wellas Informationmodeling).

Select services forprocess and UI wherepossible

Validate

Page 15: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 15

Composite Dev. - Vision

A Development Platform

That provides all the needed designand implementation facilities underone coherently integratedenvironment.

And an Execution Platform

That enables the efficient executionand effective monitoring andmanagement of the compositeapplications.

Enterpriseservices

(SAP applications)

Webservices

(Non- SAP applications)

IDE workbench

Composite applicationScript

+Specifications of

• Business objects• Business rules• Security policies, . . .

Message bus

UIframework

Scriptingengine

Security/Policymanagement

Servicebinding

Transactionmanagement

Data mapping &persistency

management

State/Processmanagement

Composite application scenario

Scriptrepository

Event/Rulemanagement

Enterpriseservices

(SAP applications)

Enterpriseservices

(SAP applications)

Webservices

(Non- SAP applications)

Webservices

(Non- SAP applications)

IDE workbench

Composite applicationScript

+Specifications of

• Business objects• Business rules• Security policies, . . .

Message busMessage bus

UIframework

Scriptingengine

Security/Policymanagement

Servicebinding

Transactionmanagement

Data mapping &persistency

management

State/Processmanagement

Composite application scenario

Scriptrepository

Event/Rulemanagement

One-stop shop for composite applications

• More concise - fewerlines of code.

• More expressive -each line does more.

• More understandable -closer to businessdomain.

• More productive -less programmingeffort required

Page 16: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 16

Value proposition

Target user = developer with SSP

Quick application implementation and extension

One coherent integrated model

On-demand and on the fly enhancement

Design-time + run-time

Incremental development: start with the domain that you prefer

Page 17: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 17

Features

DSL + Scripting = UnificationDeclarative + Imperative = GoodnessDeclarative only goes so far – we need a programming languageDSL + Ruby = Ruby DSL

Clean separation of UI, business logic and data models.

Three Kinds of CompositionProcess flow-based compositionEvent-driven compositionData integration/aggregation

Page 18: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 18

Challenges

What programming model is needed? Coarse grained or finegrained?

What is the right level of abstraction? What should be theassociated models? For UI, Security, Event Management?

From high level modeling to workflow execution and monitoring?

Page 19: Composite Dev. Research at SAP - Stanford Universitylogic.stanford.edu/.../sap_research_in_composite_dev.pdf · 2007-05-04 · Typical Composite Dev. cycle End User (LOB) 1.Define

SAP AG 2006, Novartis EBC / Wesley Mukai / 19

Technologies

Model-Based Development Environment: Integrated Development &Execution Environment; Composite Application Development;Flexible and Expressive Scripting Language to support Modeling allFunctional and Operational Aspects; Separation of UI, BusinessLogic, and Data Management; Modeling Real Time Input; AutomaticCode Generation;

Agile Execution Platform: Data Mapping and Mediation; PersistentData Management; Complex Business object support; BusinessProcess Automation; Business Process Monitoring; BusinessProcess Enforcement; Event Management & Correlations; SecurityPolicy Enforcement

Service Binding

New Service Creation