Effective Reuse From consistent requirements to variant ... · > Variant Management > Test Data...

V1.0 | 2017-01-01

Christof Ebert, Silke Riegraf, VectorMartin Bobert, Stihl

Effective ReuseFrom consistent requirements to variant management

© 2017. Vector Consulting Services GmbH. All rights reserved. Any distribution or copying is subject to prior written approval by Vector. V1.0 | 2017-01-01

MotivationIntelligent ReuseChange Project

2/29


STIHL is a globally leading manufacturer of chain saws and other power tools for professional forestry and agriculture as well as for garden and landscape maintenance, the construction sector and the demanding private user.

Welcome

3

STIHL – the world’s best-selling chain-saw brand since 1971

3/29


Welcome

4

SawsLawn mowers

Brushcutters

Hedge trimmers

Garden shredders

Blowers

Cut-off machines

PRODUCT GROUPS >>

WelcomeSTIHL – Products

4/29


…supports clients worldwide

…improves product development and product strategy, IT and provides interim management

…is as Vector Group globally present with 1750 employees and well over 400 Mio. € sales

…offers with the Vector Group a portfolio of tools, software components and services

…is growing and thus continuously hiring

www.vector.com/consulting

Vector Consulting ServicesWelcome

Railway

IT & Finance

Automotive

Aerospace

DigitalTransformation

Medical

5/29


Business Challenges – Vector Client SurveyMotivation

Cost remains biggest short-term challenge across all industries. Connectivity and Digital Transformation evolved as a major challenge.

Innovative Products

Others

Connectivity

Distributed Development

Efficiencyand Cost

Digital Transformation

Governance and Compliance

ComplexityManagement

Securityand Safety

0%

10%

20%

30%

40%

50%

60%

70%

0% 10% 20% 30% 40% 50% 60% 70%

Mid

-ter

m c

halle

nges

Short-term challenges

Vector Client Survey 2017. Details: www.vector.com/trends. Horizontal axis shows short-term challenges; vertical axis shows mid-term challenges. Sum > 100% due to 3 answers per question. Strong validity with >4% response rate of 1500 recipients from different industries worldwide.

6/29



7/29


Intelligent Reuse: Beyond EngineeringIntelligent Reuse

Strategy Concept EvolutionIntegration, certification

Development

Intelligent Reuse

Opportunistic Reuse

Intelligent Reuse implies a strategic focus across the life-cycle connecting product management, engineering and operations

8/29


Four Generic Strategies are PursuedIntelligent Reuse

Platform based products 45%

Same platform with individual modules for customization

Individualized products 15%

High degree of individualization

Standardized products 15%

Standardized products for different markets

Module based products 25%

One module set with different platforms for customization

One Platforms Several

One

Mod

ule

set

sSev

eral

Note: Percentages provide desired usage from current surveys Vector with clients.

9/29


PLE for software applications Intelligent Reuse for systems

Software-driven foundations and principles

Need translation to product culture with components and modularization

Savings achieved at software applications with many variants

Savings achieved with focus on testing, documentation, certification

Many strictly defined principles; but often dogmatically applied

PLE methodology needs adjustment to specific context and markets

Handled inside the software teams, often combined with Scrum

Driven by product management, engineering, sales

UML as lingua franca; everything is modeled

Top-down modeling to manage dependencies and variation

Focus on mere software projects and limited scope

Products with long life-cycle and high criticality, e.g. safety, maintainability

Typically applied in rather small software teams

Distributed development with suppliers and collaborative design

Introduction easy within 1 year – for application software

Introduction needs thorough changemanagement

A Practical View on PLE and Intelligent ReuseIntelligent Reuse

10/29


Over 50% of reuse programs fail Initial cost savings not

achieved No black box reuse

Reasons Product management and

sales not involved Change is underestimated and

handled ad-hoc Insufficient know-how Inadequate method and tools

Change to Reuse often FailsIntelligent Reuse

Sources: Vector Consulting Services 2012

48%

Internalresistance(e.g., interfaces)

Insufficient know-how

Pressurewithoutownership

Other reasons

22%

17%

13%

11/29



Change Project

12/29


Complex Systems: Initial Alignment StepsChange Project

StrategyProject 1

concept, bidProject 2

concept, bidIntegration, Validation

Development

Standard FeaturesCore features for all markets which evolve in a single track. Intelligent

variation management for requirements, automatic testing,

documentation, etc.

Evolution OptionsFeatures, which are currently not available

Configurable Features

Features that do not apply in all markets. Variation management

maintained on as-needed base. Test-driven requirements.

Com

pon

ents

Cu

stom

ized

fea

ture

s

Road

map

, Arc

hite

ctur

e, G

over

nanc

eBus

ines

s ca

se:

shor

t-te

rm,

long

-ter

m

13/29


Model-based Development from Architecture Design to Series-Production Readiness

Series-Production Use Cases> Requirements Engineering> Function, System and Component

Design> Software & System Design> Network Communication Design> Variant Management> Test Data Management> Change and Release Management> Design of Safety Relevant Systems

Usability and Performance> Many Users> Sporadic Users

Integration > Requested Import and Export Filters

Method and Tool SupportChange Project

Architecture Design

Series-Production Development

Requirements Design Test

14/29


Change Project

15

time

Injection 64 KByte

128

64

32

16

8

4

2

1

Cod

e S

ize

[KB

]

Increasing functionality. Increasing code size Much higher effort for testing.

FS100 first digital Igniter2 KByte

MS 280 I 1. Generation Engine management 4 KByte

WelcomeTrend ECUs for Small Combustion Engines

MS441 M‐Tronic 40 KByte

© ANDREAS STIHL AG & Co. KG15/29


Change Project

1

Structure Exact description of each single function

Discrete numbering of requirements

Complete parameter list (800 Parameters)

Versioned document with change log and history

Size: ~180 pages

Advantage Completeness with no room for interpretation

Basis for error free implementation

Mapping between RQM and test (full traceability)Autor 2/XXX

Project Manager: 2/YYY Dr. Mustermann

Stihl Project: XXXX STIHL M-Tronic

WelcomeSoftware Specification and Test – Starting point



Change ProjectWelcomeManaging Software Variants

© ANDREAS STIHL AG & Co. KG

Baseline

Harmonisation

New Project Branch

MS 2

MS 1MS 1

FS 1 Merge

It‘s difficult to manage multiple projects by Word and Excel !

17/29


Change Project

1

WelcomeDevelopment process by PREEvision

RQM

SW-specification

Project Specification

(Reuse from other projects)

Check-out models

Check-in models

project start

Found SW-Bugs

Accepted Requirement Amendments Found Specification Errors

Test Scripts according to Test Planning

Project-Specification

Testcase Creeation

Testcase Run Results

Testcasescripts

Software Development Tools

(modelbased SW-development)

CHM

Supplier

Specification

(Automated) Test Stands (HIL)

TDM

Source Control, Software Bug-Tracking

change request



The baseline contains approved requirements of all variants

New Projects will initially reuse requirements from the baseline

Requirements are structured in packages by functions Custom metric „Create Reuse Of Structure“

Copy complete structure (elements of meta class requirement package) into the projects product line

Create „reuse“ of each requirement in the selected structure

Change ProjectWelcomeReusing existing requirements…

Baseline

Project 1

Project 2

Project 3

Baseline

Resuse in Projects

reuse in different projects from one baseline Implementation in PREEvision

Metric „Create Reuse of Structure“



Project specific changes require a branch from the baseline

Reused requirements will be detached from baseline thisprocess is implemented via a custom metric

Custom metric „Detach Reuse“ = Branch-Metric

Copies requirement which needs to be branched

Removes the reuse of the baseline requirement

Creates a requirement link between the baseline and the branched requirement (the copy from step 1) and saves the version (# of check-ins) of the baseline in the link description

Change Project

excerpt of the metric‘s Java codeThe reused project after execution of „Detach Reuse“ metric

Welcome…branch…



Standardization of made requirement changes (in branches) in certain time intervals merge into new version of the baseline

Metric „Merge Branch into original requirement“

Copy branched requirement with all attributes into the according baseline requirement

Delete branch requirement link

Create a reuse of the new version of the baseline requirement in the project

Delete branched requirement

Change ProjectWelcome…and merge

© ANDREAS STIHL AG & Co. KG

Project 3

Project 2

Project 1

Baseline

Standardization

Version 1 Version 2

21/29


Implemented report generator for the project data sheets (uses calculation and diagram metrics)

Change Project

project data sheet report generator

Metric„Generate Diagrams“

Metric„Calculate Values“

Schematic structure of project data sheet report generator and screenshots of .pdf export

WelcomeData-based requirement management



Create and manage Test case specifications (Test Case Description)

Test case implementation (scripts, manual execution steps)

Test run evaluation (test coverage, planning, statistics)

Mapping between test-specifications, requirements and parameters

Change ProjectWelcomeTest Data Management

Connection between RQM and TDM

Requirements Specifications

Implementations

Analysis

Parameters

TDMRQM



Customizing for TDM

Generate config files for test cases to parameterize scripts of automated test cases for HIL-testing

Implement test case analysis on basis of test cases and whole SW-modules

Calculate degree of maturity based on planned test phases and test case run results

Change Project

Degree of maturity over timeOverview over test case analysis

WelcomeTest Data Management



Categories Challenges Techniques

Organi-zation

Ownership of assets unclear

Structure does notmatch product needs

Strengthen coherent product strategy and product-driven organization

Optimize product management, engineering and integration organizations

Culture Reluctance to invest for future

Feedback from market to engineering fails

Too much project focus

Commit clear reuse targets, related to market, revenues, cost, technical debt

Create conscious product culture Drive projects from products, not vice versa Perform business case for new features:

short-term vs. long-tern cost/benefitsEngi-neering

Complexity will grow Products needs new

features Too many project-

driven requests will erode architecture and reusability

Install proactive market-driven reuse connected to development, test,documentation, certification

Establish strong change board to judge tradeoffs based on overall business case

Maintain product and components roadmap as master for all changes

Rigorous two-way traceability Establish consistent tool support

Contractorandsupplier

Low incentive for maintaining reuse

Less control

Motivate and reward conformance to product-line specifications

Include source code in supplier agreements

The Need for ChangeChange Project

25/29


Strategic Level Align engineering, product management and sales on targets and

implementation approach Perform rigorous reviews on roadmap, change requests, feature

prioritization, market communication, dependency management, etc. Provide a “change map” with timeline, dependencies (people, projects,

markets, platforms) and strategic relevance Establish organization to align stakeholders in product management,

engineering, sales Apply strict governance criteria, namely short-term (cash flow from

projects) vs. long-term (reuse of product)

Deployment Level Start with few components which deliver results Consider activities with high savings potential, e.g., validation, automatic

regression testing, certification, component-based documentation Manage the change process with specific targets, ownership,

communication, periodic reviews, share best practices Update processes and tools along the way

Reuse Project: Strategic Change and DeploymentChange Project

26/29


Intelligent reuse helps reducing cost and improving efficiency Modularization with tailored reuse concepts Cost savings levers: automatic regression test, documentation,

certification, feature evolution, service-oriented business models

Success factors Stepwise approach starting with requirements engineering Thorough management buy-in and leadership Strong product management with transparent product strategy Architecture competence and governance

Major risks Unclear responsibilities and lack of ownership Insufficient change management, and thus fallback to the past

ConclusionsChange Project

27/29


Further ReadingChange Project

Requirements Engineering

Christof Ebertdpunkt.verlag450 pages5. fully reworked edition 2014Chinese edition, 2013

„The classic for systematically handling requirements. Written by a practitioner for practice – easy to understand and to apply! During a joint project I experienced first hand that the author knows what he talks about."Hans Leibbrand, COO, Thales

28/29


Thank you for your attention.Contact us – We are glad to support you.

Passion. Partner. Value.

Vector Consulting Services

Phone +49 711 80670-0 www.vector.com/consultingFax +49 711 80670-444 [email protected]

Effective Reuse From consistent requirements to variant ... · > Variant Management > Test Data...

Documents

Transcript of Effective Reuse From consistent requirements to variant ... · > Variant Management > Test Data...