Ontology-based Design Environment Integration (AutoCRC)

This project intends to establish a high level description of the design process in the automotive industry. In particular, it aims at developing methods and approaches for unifying concepts across functionally separated design areas with the goal to enable systematic re-use of process and product knowledge. Artefacts such as design rules, vehicle platforms, templates, and parametrics exist in various forms, but a coherent strategy to enforce structured reassessment of design knowledge is absent.

Engineering data exchange languages such as STEP/EXPRESS as well as state of the art software engineering methodologies like design patterns or UML/OCL descriptions are used to capture design rationale, design constraints and design decisions. As the ultimate objective, quicker vehicle development is expected to be achieved through higher expressiveness of existing data structures.

An important aspect of this project is multidisciplinary optimisation (MDO) of the design phase. MDO creates a wealth of data that is cumbersome to re-use and navigate. In order to support decision making, methodologies and techniques are investigated to support integration, merging and navigating of these data sets.

Research activities

Our particular interest in this project is to provide findings about the industrial employment of ontologies. An ontology can be described as a specification of the key concepts in a particular application domain and the ways in which these concepts are related, written in a machine interpretable language.

Unlike a standard relational database scheme, an ontology does not just depict what the basic tables and fields are, it distinguishes between concepts and relationships and captures part of the description why certain concepts differ from each other. E. g., an ontology can describe the distinction between a design draft, a finalized design, and a currently sidelined design variant in terms of where they occur in the production process. As a result, an ontology is more than an object oriented class hierarchy or a database schema; in practical use it can be used to generate such a hierarchy.

Ontologies and other forms of knowledge representation have a long application history for the representation of design and product knowledge and are incorporated in design support and manufacturing systems in a variety of ways. In addition, ontologies support the notion of a centralised information repository for multiple design and manufacturing applications. Examples of application areas of ontologies are:

  • Checking of requirements consistency and other design properties
  • Manufacturing planning and tool selection
  • Product-feature based process planning and dependency maintenance
  • Semantically indexed design documentation
  • Intelligent support for product configuration
  • Capture of design rationale