Summary: |
| kinodb was used to: |
- Define and implement a corporate data architecture allowing the integration of numerous
incompatible legacy asset information systems.
- Support change control processes to enable the coherent management of the data
dictionaries of the legacy systems.
- Implement a generic 'universal' asset information repository that could
be used to warehouse data from other systems, enabling the retirement
of redundant data stores.
- Publish web-distributable details of corporate asset type definitions including data collection
requirements, standards and regulatory constraints.
|
Business background: |
| Key background information relating to the organisation included: |
- A regulated infrastructure maintenance organisation with a multitude of disparate data sources developed
over many years.
- Numerous 'home-grown' and tactical asset information sources developed to differing standards.
- Multiple, inconsistent definitions of key corporate asset types with resultant inability to gain
a single view of asset information.
- No central management of the data specifications of the various asset information repositories.
- Poor dissemination of definitions, standards etc.
|
Challenge: |
- Design and implement a single repository capable of recording the master definitions of asset types,
each mapped to the disparate definitions supported by the various existing asset information systems.
- Enable web-based reporting of asset
type definitions and associated data collection standards.
- Use these master asset type definitions to drive the implementation of a 'universal' asset register
capable of maintaining data from existing and future systems.
- Enable the rules-based cleansing of data in existing systems.
- Enable migration of data from existing systems to enable their retirement.
- Enable all data maintenance to be performed by business users, not IT staff.
|
Solution: |
| kinodb was used in
the design and definition of an architecture for the maintenance of
the organisation's asset information. This architecture was based
on the central maintenance of the data dictionaries of each of the
organisation's core asset information systems, allowing the asset
information recorded in each to be related to and from a standardised
set of asset type specifications. |
| Using kinodb's rapid
development capabilities, the initial solution was prototyped in user
workshops to demonstrate the validity of the proposed data model.
This initial implementation was then used as the basis for the development
of the production system. Additional workshops were conducted with
business users to determine their requirements for support of change
control processes, reporting requirements and so on. Functionality
in support of these requirements was developed incrementally with
the close involvement of the users, who were able to use early versions
of the system for data entry use before the delivery of the completed
system. kinodb's rapid development cycles meant that users could see
the results of requested changes very quickly, leading to increased
user involvement at all stages and a corresponding high degree of
user confidence and ownership of the implemented solution. |
Techniques: |
| This project made extensive use of kinodb's capabilities: |
- Automatic creation of all user input forms. In a relatively complex system (with dozens of major
entities) the design and management of user input forms is a major task, and makes the iterative development of systems prohibitively
time-consuming and error-prone. It is also expensive as each change requires reworking of previously developed input forms. kinodb dynamically generates each input forms as it is required.
- kinodb's SQL Generation Engine removed any need for the coding of complex SQL statements. Again, this
is a major enabler of iterative systems development as changes of any magnitude can quickly be made without the need to recode
and retest thousands of lines of SQL code
- Rigorous maintenance
of data normalisation. Any complex data design is compromised if data is denormalised. kinodb's ability to transparently manage
the referential integrity of complex relational data sets allows the designer to concentrate on the design of the database without
being concerned with the complexity of the SQL that will be required to implement and manage it.
- Automated maintenance of hierarchies of types was used, enabling the inheritance of
asset type characteristics (attributes, relationships etc.) down the hierarchy of asset types. The ability to transparently maintain
these hierarchies enables the implementation of a relational database that exhibits characteristics more normally associated with object-oriented
technologies.
- Fast generation of HTML reports for web deployment. As kinodb fully understands
the structure of the database, creating complex reports from the data model is simple.
- Support for Binary Large Objects (BLOBS). This allowed the storage
of data collection guidance diagrams and photographs of sample
assets.
|
