ftp.umcs.maine.edu/2015-02-07.ftp.umcs.maine.edu.tar/ftp.umcs.maine.edu/pub/WISR/wisr6/proceedings/ascii/navarro.ascii

Text File | 1993-10-19 | 17KB | 345 lines

    Navarro - 1

    Navarro - 1


Organization Design-Based
Software Reuse Adoption Strategy


James J. Navarro

Hewlett-Packard Laboratories
1501 Page Mill Road, Palo Alto, CA  94304
Tel: [phone], fax: [phone]
Email:  [e-mail]


Abstract

Hewlett-Packard divisions piloting software reuse, as well as many non-HP
reuse efforts, have identified organizational issues as major impediments to
the success of their reuse efforts.  The design of software reuse work
processes and their integration into the larger business organization is
critical to overcoming many impediments affecting the success of software
reuse.  HP Lab's Software Reuse Department is developing a reuse adoption
strategy based on expertise and experience from the field of organization
design.   The adoption strategy's purpose is to facilitate the analysis, design
and implementation of a reuse-based software engineering organization
which we call the Flexible Software Factory.


Keywords:  Reuse, organization design, adoption strategy, change
management, organization, management, core process, flexible software
factory, software factory

Workshop Goals:  Networking; exchanging perspectives on effective
process analysis / design and organization designs for software reuse;
stimulating discussion on innovative management approaches to
organizing work; and learning

Working Groups:  Reuse management and organization; reuse process
models; and reuse adoption strategies

1   Background

Hewlett-Packard Laboratories' Software Reuse Department is involved in multi-
disciplinary research focussing on the problems inherent in implementing
software reuse.  We are exploring two primary dimensions of research within
reuse:  the business /   organizational design dimension, and the supporting
methods, tools and technologies dimension.  These dimensions are represented
in our research on the Flexible Software Factory and Domain Specific Kits.  See
Martin Griss's WISR '92 position paper for more details on our department's
research [5].  This paper is an update on our research described in my WISR'92
position paper [6], and will focus primarily on the business / organizational
design dimension of our department's work.

The author is a Principal Project Scientist / Engineer in Hewlett-Packard
Laboratories.  He has been involved in several successful organization design
efforts in HP divisions creating high performance organizations for product
development and manufacturing.  He currently is leading a reuse-based
process modeling / analysis pilot in one of HP's Divisions implementing a reuse
oriented software development strategy.  His role in the Software Reuse
Department is to provide technical vision and leadership in the research effort
on the Flexible Software Factory, [5] and [6] .

2  Position

HP divisions who are piloting reuse, as well as other organizations like GTE,
AT&T, and IBM, have identified organizational issues as major impediments to
the success of their reuse efforts.  A few examples of these impediments are:

Ñ      software reuse producers are 'borrowed' in order to solve short-term
programming needs,

Ñ     individual and organization incentive systems are not supportive of the
long and short-term viability of software reuse and its producer / consumer
roles,

Ñ    the time and resource investment required to benefit by software reuse goes
against the shorter-term, results-oriented focus of managers, and

Ñ     programmers are resistant to reuse because they perceive that their
creativity will be limited.

The design of the software reuse organization is critical to overcoming many of
these types of 'non-technical' impediments affecting the success of software
reuse[3].  The Flexible Software Factory (FSF) research within HP Lab's Software
Reuse Department is addressing these and other key 'non-technical' issues by
applying successful organization design methods and strategies to reuse.  Our
approach includes the use of an adoption strategy to help initiate and guide
change from a non-reuse based software development organization to a reuse
based software development organization.  The outcome of this adoption
strategy is to create a software development organization that instantiates and
aligns work and work processes, and organizational resources, goals, objectives
and intentions to support software reuse.  In addition, the adoption strategy
can be used to improve existing reuse practices.  This position paper describes
an adoption strategy which we call the Flexible Software Factory (FSF) Adoption
Strategy.

We are in the process of applying and validating portions of the FSF Adoption
Strategy in a HP Division.



2.1  The FSF Adoption Strategy

The FSF Adoption Strategy consists of a series of phases and activities whose
purpose is to develop a systemic understanding of a software engineering
organization in order to determine what changes are needed to effectively
infuse reuse into its production process.  This systemic understanding requires
comprehension of the organization's production processes, technologies,
organization and people, and the interplay between them.  The FSF Adoption
Strategy discussed in this paper is a methodology for transforming a software
engineering organization from its current state to a new steady-state where
reuse is instrumental to the software production process.  In addition, the FSF
Adoption Strategy can be used to improve existing reuse-based software
development organizations.

The FSF Adoption Strategy consists of four phases.  These phases are
represented in Figure 1.1.


Figure 1.1  The Phases of the FSF Adoption Strategy

The following sections are introductions to the first three phases of the
adoption strategy.  They will cover the purpose, the top level questions to be
answered, and the deliverables of each phase.  In addition, they will describe
the activities to be performed in each phase.


2.2 FSF Preliminary Assessment and Readiness Phase

The purpose of the preliminary assessment and readiness phase is to
determine if there is an opportunity for reuse and if so delineating the
characteristics and value of that opportunity; to define the boundaries and
primary purpose of the software development organization (which we'll refer to
as "Organization" throughout the document) in order to create a shared
understanding and support among the appropriate stakeholders both inside
and outside the Organization, and finally, to develop investment criteria to
guide the later phases of the FSF adoption strategy.

There are two top level questions to be answered in this phase.  They are:
What is the reuse opportunity, and the potential benefits, costs and risks
associated with the reuse opportunity?  What is the purpose, context, and
boundaries of the Organization?

The outcomes of this phase are to determine if the potential opportunity for
reuse is worth the investment of time and resources, and the possible risks to
be overcome; to judge the readiness of the business unit for change; to
approximate the cost, scale, time-frame and objectives of  a FSF; and finally, to
make a decision to continue to the comprehensive assessment and analysis
phase, and create a change management structure to support this work.

To determine the reuse opportunity and FSF investment criteria we propose
performing the following assessment activites in the Preliminary Assessment
Phase:  Domain Reuse, Market and Asset assessments.  Combined, these
assessments will begin to uncover the investment parameters of the reuse
opportunity and related risks.

In parallel, the determination of the Organization's boundaries, purpose and
readiness for change are required.  This information will help Organization
stakeholders understand the scope and effort required to change, and to begin
to obtain needed support.

See Figure 2.1 for an illustration of the activities and deliverables produced
within the Preliminary Assessment and Readiness phase.



Figure 2.1  Activities and Deliverables of the Preliminary Assessment and
Readiness Phase

2.3  FSF Comprehensive Assessment and Analysis Phase

The purpose of the FSF Comprehensive Assessment and Analsis Phase is five-
fold.  Within the scope of the FSF it is:  First, to articulate how the current
Organization is achieving the FSF's purpose, objectives, and philosophy.
Second, to establish baseline metrics within the current Organization that
relate to the FSF objectives, and performance targets and goals.  Third, to
analyze current Organization processes, elements and mechanisms.  Fourth, to
present the current Organization as a system determining gaps between
existing state and desired FSF state, identifying key drivers that if fixed will give
the biggest benefits.  Finally, to determine changes needed in current
Organization.

There are three top level question to be answered in this phase:  First, how is
the Organization achieving the FSF's business, organization, and process
requirements today?  Second, what are the Organization's baseline measures
regarding the FSF process, financial, and social objectives, and performance
targets and goals?  Finally, when viewed systemically using the FSF elements,
where are the key drivers in the Organization that if fixed will give the biggest
benefits?

The deliverable from this phase is a document focused on the current
Organization as it relates to the FSF, it includes:  detailed process maps, process
baseline metrics, process analysis, process control table, an analysis of
communication and decision-making mechanisms, an analysis of
organizational roles and structures, and a portrayal of the Organization as a
system showing areas needing to be addressed.

The outcomes of this phase are to review and adjust the FSF investment
criteria developed in the Preliminary Assessment and Readiness phase based on
the learning in this phase; to obtain agreement on the gaps between existing
and desired state, the Organization's key drivers that if fixed will give the
biggest benefits, and the changes needed in the current Organization; and
finally, to determine whether to proceed to the next phase:  FSF Design.

To describe the current state, the key drivers and changes needed in the
Organization we propose performing the following three comprehensive
assessment/analysis activities in the FSF Comprehensive Assessment and
Analysis Phase:  process, business, and organizational assessment/analysis
activities.

In addition, a systemic description of the of the current state will be used to
synthesize the seperate assessments/analysis listed above, and provide the
mechanism for identifying key drivers and needed changes in the
Organization.

See Figure 3.1 for a illustration of the activities within the FSF Comprehensive
Assessment and Analysis phase.


Figure 3.1  Activities and Deliverables of the FSF Comprehensive Assessment
and Analysis Phase

2.4  FSF Design

The assessments and analysis are done and ideas are collecting for the design
of the FSF.  We are now ready to collect the ideas and design the FSF.  How do
all the pieces fit together?  How does the design team integrate and make sense
of all the individual ideas generated during the assessment/analysis phases?
The analysis process to date can seem exhausting and painstaking.  The
analysis was specific and detailed.  It's time to stop and put systems theory to
work and begin designing.

The purpose of the FSF Design Phase is to create an effective FSF design taking
into consideration relative value and risks; to develop a successful FSF
implementation plan; and to build acceptance for the new FSF design and FSF
implementation plan by the stakeholders through participation and
communication.

There are three top level question to be answered in this phase:  First, what is
the ideal and optimal FSF design(s)?  Second, what are their relative value and
risks?  Third,  how can an implementation be devised that would both
decrease the risks and increase the success of the FSF?

The documentation produced in the FSF Design phase includes a set of FSF
design(s) with an assessment of their relative value and risks,  a provisional FSF
design, and the FSF implementation plan for the provisional FSF design.

The outcomes of this phase are two-fold:  First, through several
presentations/interactions iteratively improve and narrow the number of
optimal FSF designs leading to the choice of one provisional design.  Second, to
obtain agreement on the FSF implementation plan for the selected provisional
design.

To produce the selected provisional FSF design we propose the following
activities:  development of ideal FSF design(s), joint optimization of FSF
design(s), assessment of relative value and risk of optimized FSF design(s), and
the selection of the provisional FSF design (usually through the combination of
the best ideas expressed in several designs).

An additional activity to produce the FSF implementation plan for the
provisional FSF design will be needed once the provisional FSF design is
selected.

See Figure 4.1 for a illustration of the activities and deliverables within the FSF
Design phase.



Figure 4.1  Activities and Deliverables of the FSF Design Phase

2.5 Summary

It is believed that effective systematic reuse requires an appropriate supporting
organization structure [1,2,4,7].  From an organization design perspective there
needs to be a method for designing that structure.  That method is
instantiated in the FSF Adoption Strategy.

We see an organization's structure consisting of  a number of elements
including decision-making and problem-solving processes, management and
team roles, communication processes, measurement and feedback systems,
evaluation and rewards, and vertical and horizontal management structures.
Together these elements comprise the software development organization.  In
order to change a software development organization to a reuse-based
organization or to improve an existing reuse-based software development
organization we believe that a systematic analysis of these elements are
important, as well as a portrayal of how they interact as a 'system'.  This
portrayal will give a sobering view of why a software development organization
is achieving or not achieving its reuse objectives.  We feel that the design of the
reuse-based FSF needs to be based on the alignment of these elements.

3  Comparison

An alternative to the FSF Adoption Strategy is the Software Productivity
Consortium's (SPC) Reuse Adoption Process.  SPC's process is a technology
transfer process for transferring reuse technologies (processes, methods, and
tools) into an organization [8].  Whereas the SPC's process focuses on
transferring reuse technologies, the FSF Adoption Strategy focuses on
understanding the organization as a system, determining what technologies to
transfer and what changes to make based on this systemic interpretation.


4 References

[1]    V.R. Basili.  Software Development:  A Paradigm for the Future (Keynote
Address).  Proceedings COMPSAC '89, Orlando, FL, September 1989.

[2]    J.W. Hooper and R. Chester.  Software Reuse - Guidelines and Methods.
Plenum Press, New York, 1991.

[3]    W.B. Frakes.  An Emperical Framework for Software Reuse Research.  In
Third Annual Workshop:  Methods & Tools for Reuse.  CASE Center,
Syracuse University, June 1990.

[4]    M.L. Griss. Software Reuse Handbook.  Hewlett-Packard Corporate
Engineering, Palo Alto, CA, 1992.

[5]    M.L. Griss. A Multi-Disciplinary Software Reuse Research Program.  In
Proceedings of the 5th Annual Workshop on Software Reuse, Palo Alto, CA,
1992.

[6]    J.J. Navarro.  Organization Design for Software Reuse.  In Proceedings of the
5th Annual Workshop on Software Reuse, Palo Alto, CA, 1992.

[7]    M.A. Simos.  Software Reuse and Organizational Development.  In
Proceedings of the First International Workshop on Software Reusability,
Dortmund, 1991.

[8]    Software Productivity Consortium.  Reuse Adoption Guidebook. SPC-92051-
CMC, Hendon, VA.

5 Biography

James Navarro is a Principal Project Scientist / Engineer at Hewlett-Packard
Laboratories.  He is the 'technical lead' in a multi-disciplinary research
program focussed on applying organization design to software reuse.  He began
his research on organization design in 1988 while working in Apple Computer's
Strategic Technology Group.  While obtaining a graduate degree at Harvard
University's Graduate School of Business he performed research with professors
Rosabeth Moss Kanter and J. Richard Hackman.