"Building Intelligent Content from 30 Years of Legacy Documents," Intelligent Content 2010, Paul Wlodarczyk

Case Study: Building Intelligent Content
from 30 Years of Legacy Documents

Paul Wlodarczyk
Director, Solutions Consulting
Earley & Associates

Copyright © 2010 Earley & Associates Inc. All Rights Reserved.

COPYRIGHT © 2010 EARLEY & ASSOCIATES INC. ALL RIGHTS RESERVED.

What do we assume about Intelligent Content?

We usually assume:
– Loads of systems
– Information design
– Structured content
– Tied to the business
– Personalized
– Semantic
– Etc.

We assume new.

2

Case Study: Applied Materials
Global leader in Nanomanufacturing Technology solutions for the
fabrication of semiconductor chips, flat panel displays, solar
photovoltaic cells, flexible electronics and energy efficient glass
– Founded in 1967, $5B revenue, 8 Product Business Groups
– Produces the technology that helps produce virtually every semiconductor chip and
flat panel display in the world
– Long-lived products – some in service over 30 years
– 2500 customer engineers (“CEs” – 20% of work force) maintain, diagnose, and repair
their equipment in customer fabs worldwide
Business need:
– Improve CE troubleshooting and repair performance in the fab through improved
information access

3

What do we mean by Intelligent Content?

• In our client’s words:
– “Deliver the right information at the right time
information, time,
in the right format to the right person
format, person.”
• In our words:
right information = Relevant, accurate, up-to-date, approved
right time = In the context of work / business process
right format = Accessible
right person = Authorized

4

Information Challenge #1:
30 Years of Legacy Content
• …and lots of it
– PDFs – manuals, drawings, some are scans
– Communities – forums, shared drives, bulletin boards
– Knowledge Management – tips, Best Known Methods
• How much gets “chunked”? How to convert it?
• How to search it effectively?

5

Point of need is in the clean room
• Physical Access for the CE
– Double / triple gloves
• Typing/pointing are difficult
– No place to set a laptop down
• Wireless / Internet Access
– Closest point of access to company
internet may be across town
• Customer Information Security
Concerns
– Protection of trade secrets
– Some won’t allow laptops in fab
– Some won’t even allow paper in fab
• “Tools” are all unique in the fab
– Need docs that reflect product “DNA”

6

KM is complicated by IP / InfoSec concerns
• Garden-walling Customer IP
– Have to be sure that when a CE
shares a tip or asks for help that they
are not exposing customer IP to other
customers (with whom they compete)
• Protecting Company Product IP
– DRM for preventing piracy / reverse
engineering of products, parts
• Preserving Company Services IP
– Assure that customers pay for value-
added services
– Secure service IP from competitive
service providers

7

An “archipelago” of content centric systems

PLM
Tech Pubs Library
Product
Library

Forum
Product
Product Library

Drawings DRM Forum Library
Forum
Forum

Product Product
Product Forum
Library Library
Library
Forum
Email
SAP Parts
Diagnostics
QA Product
Library

Best Practices
Search and integration are issues Product
Library

8

Where to begin eating the elephant?
• Key observation:
“Last thing we need is another system”
• Need a balanced viewpoint –
the “Four Pillars of KM”:
– People
– Process
– Technology
– Content
• Assess the current state
• Define the desired state
• Define the gap closing actions

9

Technology
• Technical Publishing
– Authoring: Office, FrameMaker
– Team Center - Drawings
– Publishing in PDF, Live Cycle DRM
• ECM
– LiveLink  SharePoint
• ERP
– SAP for service billing, parts ordering, etc.
• Various proprietary systems
– Parts, diagnostics, KM, etc.
• Connectivity and access to laptops in the fab
– Ranges from “none” to “near” to “Nirvana”
– Need to support a low-tech solution

10

People
• Several key user groups:
– CEs (content consumers and contributors)
– Support (content consumers and contributors)
– Tech pubs and training (content publishers)
– Engineering teams (content contributors)
• Understood roles, expectations,
alignment, barriers to change
• Key findings:
– Content authors had well-defined needs and
were extremely engaged
– Content publishers had good alignment for
change and good orientation around topic-
oriented content
– Strong need for governance over technology
projects

11

Process
• Service Work / Business Process
– CEs spend lot of time searching – mostly
during diagnosis
– Business value is not in reducing search time; it
is reducing product downtime.
• Content Lifecycle
– Authoring, Publishing processes fairly traditional
– More engineering contribution and editing than
authoring
• Collaboration
– KM starting to hit its stride. Sorting the IP issues
• Program Governance
– Service IT programs were stove-piped
– Coordination needed for true intelligent content

12

Content

• Monolithic PDF needs to be “chunked”
• Terabytes of content
– Most relevant content is PDF, tagged to
corporate taxonomy, in a CMS
– Some scans of 30 year-old manuals
• Sales-oriented Taxonomy
– Needs to reflect the service view (failure
modes, symptoms, function, etc.)
– Taxonomy governance process neglected
• Info Security
– Access Management and DRM is essential

13

Roadmap to Intelligent Content

Phase 1: Phase 2: Phase 3:
Foundations for Access at Point of Intelligent Content
Intelligent Content Need Platform
• Service Technology • CE Platform for Fab • Common client platform
Council Access • SOA-based application
• ECM implementation • Intelligent work packages integration
• Faceted search • CE training, certification • Interactive Electronic
• Federated search • XML Pilots Technical Manual (IETM)
• Taxonomy development viewer
• Modular content, tagged • XML content
• Continuous Improvement • “Smart art”
for other service systems

14

Project deployment
• Service Technology Governance
• Modular Content Strategy
• Taxonomy and Metadata Development
Five Core • CMS IA and User Experience
Activities • Search and UX Technology

• Several divisions, several products
• Real users, active service data, “live fire” trial
Production • Collect data, evaluate, and refine
Pilot

• Plan for what to migrate and how
Scale and • Get to an 80%-ish solution
Migrate

15

What about XML?

• Can publish component PDF without XML
– Largely maintain components at the source level
today in Word and FrameMaker
• Chunking and tagging legacy content will inform a
content model for XML
• No near-term drive for XML
– No real reuse – little “transcluded” content
– No multi-channel publishing
– No localization
• Need for XML in 3-5 years
– Emerging need for localized content
– Need to integrate XML content into IT systems
for real service dashboard – dynamic publishing

16

Intelligent Content Myths

• You need XML for intelligent content.
– We can get there with well indexed component
content. The key is relevant semantic markup.
• We can just find everything with enterprise
search / Google.
– Taxonomy and component content strategy are essential
for relevancy and contextual navigation.
• We can justify the project with the time
we save by not searching.
– In the real world, projects can only get sponsored with a
real ROI at every phase, and buy-in of all stakeholders.
• Intelligent Content is futuristic.
– We can get there TODAY.

17

Advantages of the “four pillars” approach

• Comprehensive view of the
issues, benefits, and
constraints
• Achieved buy-in at all
levels of the organization
• Yielded a clear roadmap to
Intelligent Content
• Clarified potential phasing
of a large, complex project

18

Questions/Discussion
Paul Wlodarczyk
Earley & Associates, Inc
585-598-6050
paul@earley.com
www.earley.com


About Earley & Associates, Inc.

• Focus: Information Architecture • Consulting Philosophy:
(“IA”) Services • Organizing Principles based on
business context and goals
• Founded: 1994
• Four Pillars - People, Content,
• Personnel: Twenty core team Process, and Technology
consultants, plus a network of
other top industry experts
• ECM and KM experts
• taxonomy specialists
• search experts
• information architects
• usability professionals
• technology consultants
• business process experts
• Headquarters: Boston, MA

20

"Building Intelligent Content from 30 Years of Legacy Documents," Intelligent Content 2010, Paul Wlodarczyk

More Related Content

What's hot

What's hot (10)

Similar to "Building Intelligent Content from 30 Years of Legacy Documents," Intelligent Content 2010, Paul Wlodarczyk

Similar to "Building Intelligent Content from 30 Years of Legacy Documents," Intelligent Content 2010, Paul Wlodarczyk (20)

More from Earley Information Science

More from Earley Information Science (20)

Recently uploaded

Recently uploaded (20)

"Building Intelligent Content from 30 Years of Legacy Documents," Intelligent Content 2010, Paul Wlodarczyk