SGML &, schemas: from SGML DTDs to XML-DATA.   Table of contents   Indexes   XML and the ATA Interchange Model
 
 

Realising the Potential of Object Technology Through New Working Practices
, Experience and Innovation at Ford of Europe
 
John   Chelsom
  Managing Director
  CSW Informatics Ltd
Oxford Centre For Innovation
Mill Street
Oxford   United Kingdom  OX2 0JX
Phone: +44 1865 794 789
Fax: +44 1865 205 008
Email: john@csw.co.uk Web: www.csw.co.uk
 
Biographical notice:
 
John Chelsom
Neutral
Technical Support Operations, Ford of Europe
Ward, Philip
 
John Chelsom is Managing Director of CSW Informatics, a company dedicated to providing object-level information management solutions using SGML and database technology. Originally trained as an electrical engineer, John worked first as an X-Ray engineer and later gained a PhD for work on the application of knowledge based systems in medicine. From there it was a short step to the world of SGML where he has been responsible for the design and development of SGML information management systems using both object and relational database technology. John is also the presenter of the Technology Appraisals seminar series on SGML Document Databases.
 
Philip   Ward
  Publishing Services and Compliance Supervisor
  Technical Support Operations, Ford of Europe
Room 20/750
Arisdale Avenue
South Ockendon
 Essex  Neutral  RM15 5JT
Phone: +44 1708 858 384
Email: pward5@ford.com
 
Biographical notice:
 
Philip Ward
 
Philip joined Ford as an apprentice and went on to work as a technical author, before moving into the systems side of technical publications. He has played a major part in Ford's globalisation initiatives for technical service information and now leads the team responsible for the systems architecture and implementation of the next generation of publishing systems in Europe. Along the way he has gained an MSc in Distributed Information Systems and found time to indulge his passion for motor bikes, climbing and roller blades.
 
ABSTRACT:
 CORBA, Common Object Request Broker Architecture 
 JAVA 
 business objects 
 components 
 
For several years, Ford of Europe have been producing and delivering technical service information using SGML. Ford are now developing innovative solutions to managing SGML components over global networks by combining SGML/XML with cutting edge technologies such as JAVA, CORBA and Business Objects. SGML and object technology have been used to seek solutions to problems such as information reuse, standardisation of terminology, distributed authoring and translation management. The challenge for Ford is in introducing ground-breaking technology and new working practices within the framework of an on-going SGML production environment.
 
 

Introduction
 
As part of Ford Motor Company, Technical Support Operations - Europe (TSO-E) is responsible for all technical service information to accompany European-engineered vehicles world wide. Publications produced range from Workshop Diagnostic Manuals to glove box Owner Guides and information is delivered on paper, fiche, CD ROM or via the web. To support this operation, a business systems group exists within TSO-E . This group is responsible for the technology used to create and distribute the information to the end users.
 
In recent years significant new challenges have had to be tacked including an increase in off-site authoring, shorter publication lead-times and increased reliance on electronic delivery of information. To meet these challenges, new processes and technology has been designed to enable technical information publishing in a distributed enterprise environment.
 
 

Business Drivers for Change
 
There are a number of very important business goals driving the change to working practice within TSO-E, which are typical of most global high technology manufacturers:
  • make electronic information the primary delivery medium
  • change business processes to support Globalisation
  • improve the quality of information
  • reduce the costs of production
  • shorten the time to market
 
Current production processes have evolved for the delivery of paper and microfiche information. The introduction of Ford's Technical Information System (TIS), which uses CD to deliver technical service information to dealerships, has brought new challenges. Over 100,000 information units must be managed on a single CD. Due to the complexity and quantity of information Ford can no longer rely on manual processing to manage this data.
 
The move towards globalisation has brought with it new challenges for both Ford and the outside agencies that provide technical authoring and translation services. During last year these agencies have moved from being supplemental services to become complete, autonomous solution providers for Ford. The challenge has been to ensure that quality standards are maintained and that information can flow easily between Ford and the agencies.
 
The adoption of an SGML-centric approach for some publications has now positioned Ford to mechanise more; therefore reducing costs and time-scales. However, to implement this approach across all publications, it is vital to eliminate any post-processing of information delivered by the agencies and ensure that this information is of the highest quality.
 
 

Changes to Working Practice
 
Initially, process change is concentrated on Workshop Manuals and Technical Service Bulletins. Publication updates or the generation of new vehicle information will be controlled through a Product Specification. The Product Specification is vehicle specific and will provide a detailed list of all the procedures that require either creation or update. It will give a visual indication of the status of information during authoring and translation, throughout the creation of a publication. In addition, it will provide planning; co-ordination and detailed tracking information for use at authoring review. The product specification will be built using controlled procedure titles and labour operation codes.
 
Eventually, the requirement will be that all agencies will create information using a native SGML editor. This will allow a lower cost, PC-based authoring solution and provide improved support for SGML.
 
Many agencies have been working for some time installing or improving telecommunications links with Ford. To use the new processes some steps can only be achieved interactively using a web browser. Therefore, connection to Ford using the recommended telecommunications links will be mandatory in the future.
 
The translation process will also be affected. Information will be passed automatically to the external translation agency by the system rather than manually by the authoring agencies. Also, for Workshop Manuals and Technical Service Bulletins, the translation agency will be required to work with SGML.
 
Composition of SGML information into a printable form will be completed by machine and controlled by the system. All languages will be composed in this way — authors and translators will be concerned only with the SGML information, not with page composition and layout.
 
 

Benefits of New Practices
 
The forthcoming changes will impact on each stage of the information creation cycle, from planning to delivery. By affecting a change over the entire process, maximum benefit will be gained, including:
  • Provide status information on each piece of data held within the system, which will be available on-line and on-demand
  • Minimise manual tasks and remove manual post-processing of data
  • Streamline the translation process
  • Provide immediate access to language versions of publications
  • Introduce automated methods to ensure data quality and conformity to standards, including terminology
  • Provide data management tools to control processes
  • Allow the use of cheaper PC based authoring tools
 
These changes will make it significantly easier to publish information on TIS as well as reduce cost, improve quality and reduce time to market.
 
 

New Technology and System Architecture
 
Underpinning the process changes is a new systems architecture for managing the SGML source information and handling distributed production processes. At the heart of the system lies an document repository, to which agencies gain access through web browsers on an Intranet. The key technology that enables fully distributed processing a three-tier client-server architecture, using business objects and implemented using JAVA and (in the future) a CORBA compliant object request broker.

 
System Architecture Using Business Objects

 
 
 

Leveraging Object Technology
 
At present SGML information is managed at a relatively high level of granularity — the service procedure level. By moving towards a fully SGML production process using business objects, TSO-E are positioning to gain advantage from management of SGML components at a lower level in the document structure.
 
Several years ago, TSO-E developed a plug-in module to the authoring environment which helped technical authors standardise the terminology used service publications. It had been discovered that as much as 70% of the SGML text in a Workshop Manual was in fact a repetition of text from a relatively small set of (several thousand) service procedure steps. In order to reduce the amount of typing for the technical authors, and to ensure that authors used a common set of terminology, the service procedure steps were loaded into a relational database so that they would be accessible to all authors. The key to making the authors reuse the stored database text, rather than retype it, was to find an easy method of accessing with the database while using an SGML editor. The method found by TSO-E was very simple: as the authors typed the text of a procedure step in their SGML editor they could press the carriage return key, which would kick off a search for steps stored in the database matching the text they had typed so far. If there was a single matching step then this was inserted into the SGML document, replacing the text typed by the author; if there were several matches in the database then these were presented to the author in a selection box.
hot-key reuse
 
This basic idea of hot-key search, pioneered by TSO-E using a relational database, has since been extended in other applications to provide hot-key reuse from an object database. With SGML documents stored at component level in the object database there is the opportunity to reuse small components (such as Ford's procedure steps) between SGML documents. The benefits are obvious — with as much as 70% of any document being reused from the small set of SGML elements, the cost of storing, maintaining and (most importantly for operations such as Ford's) translating documents can be dramatically reduced.
 
Hot-key reuse using an object database works in the same way as Ford's database look up, but when a search hit is found, it is inserted into the document as a reuse of the SGML element found in the database. This provides users with an easy method for creating documents containing reused text at the fine-grained level of the document hierarchy, which is one of the key benefits to be gained from storing complete SGML groves in a database. Without hot-key reuse, this has always been difficult to achieve in terms of the user interface, because there has been no easy way for users to select the correct element to insert into a document, when there may be many thousands of choices of small sized SGML elements.
 
At Ford, the first step in implementing technology for controlled terminology on a wider scale is to develop the controlled terminology data set, through careful linguistic analysis of existing publications. Once this work is completed, it will be possible to leverage the power of the distributed object architecture to allow all agencies to use the controlled terminology. There are other advantages to using controlled terminology, over and above the simple reuse of text:
  • Documents are more readable in the source language, because there are less ambiguities
  • Text is easier for non-native readers to understand
  • Better search matches are found when using translation tools such as authoring and translation memory
SGML &, schemas: from SGML DTDs to XML-DATA.   Table of contents   Indexes   XML and the ATA Interchange Model