CREATOR

 
Main Page

Architecture 

Case Representation 

Memory Structure 

Problem Elaboration 

Case Retrieval

Case Representation
Design episodes are represented in the form of SBF models. These models are based on the component-substance ontology developed by Bylander and Chandrasekaran. A case is composed by three parts : problem specification, solution and explanation. The problem specification comprises a set of high level functionalities (HLF) and functional specifications (FS) that the new design must have. The case solution describes the design structures. The explanation is represented in terms of sequences of  transitions between behavioral states.

High level functionalities are abstract and created to help the user specifying the design problem. A design problem is a tree of functionalities, where leaves are functional specifications and the other nodes in the tree represent high level functionalities. This taxonomy of functionalities makes the task of specifying the design requirements more flexible and easy to perform. A functional specification is a schema that represents the initial behavior state, the final behavior state, a pointer to the internal causal behavior of the design, inputs from the external environment and structural constraints. 

The design solution is a partonomic hierarchy of device structures. Each structure can be viewed as a set of connected sub-structures. Substances can flow through structures. The structure schema is as follows :

  • is-a-part-of : the super structure of the structure;
  • consists-of : the sub-structures of the structure;
  • is-a : the class of the structure;
  • relations : structural relations with other structures;
  • properties : structure properties;
  • functions : the structure primitive functions;
 A case explanation is the causal behavior of the design in terms of as directed acyclic graphs (DAGs). The nodes of a DAG represent behavioral states and the edges represent state transitions. A behavioral state can be composed of one or more substance schema’s. A substance schema characterizes the location, properties, and property values of a substance. A state transition schema has the following slots :
  • using-functions : The primitive component’s functions that allow the state transition to happen;
  • under-condition-relation : structural relations needed to make the state transition to happen;
  • domain-principle : domain principles or laws responsible for the state transition;
  • under-condition-component : component conditions that must be met;
  • under-condition-substance : substance conditions that must be met;
  • under-condition-transition : other state transitions that are influential to the state transition.