TY - JOUR
T1 - Schema-based conversation modeling for agent-oriented manufacturing systems
AU - Lin, Fuhua
AU - Norrie, Douglas H.
N1 - Funding Information:
We have proposed a new approach to modeling conversations, for agent communication and cooperation. The concept of conversation schema links tasks and task-related information (topics) with low-level communicative acts. This approach has the following distinctive advantages. First, it ensures consistency and effectiveness of agent conversation by considering sub-task constraints. Second, it lowers communication overheads by incorporating CMs that can quickly determine for the participating agents what to do without resorting to lengthy reasoning by them. Third, it reduces complexity of implementation by constructing CMs that separate the description of common agents’ functionality from that of communication and synchronization, to ensure local and global coherency. Fourth, it enhances the reusability of software components. Domain-independent and domain-specific conversation knowledge are organized and formulated into hierarchies of conversation schema classes using object-oriented methodologies. Finally, flexibility is realized by incorporating decision-making units and control mechanisms via arc expressions and tokens of nets. The initial markings of CPN representing a schema are used to represent the kinds of situations the agent will likely encounter. Therefore, they provide a possibility for incorporating learning mechanisms into conversation schemata so that CMs are capable of learning emergent relationships from past agent interactions, for near optimization and rapid response during task–schema reasoning. Such learning mechanisms are included in proposed future research. Also as a future research topic, is building security and trust mechanisms into schema-based conversation modeling, to realize the level of secure agent communication that is particularly important in Web-based manufacturing applications. Dr. Fuhua Lin is Associate Professor of Center for Computing and Information Systems at Athabasca University in Canada. Before joining Athabasca University, he was working at Intelligent System Group (ISG) at University of Calgary as a postdoctoral research fellow and the National Research Council of Canada as an Assistant Research Officer. He obtained PhD from Hong Kong University of Science and Technology (HKUST) in 1998. He was involved in many research projects on the development of intelligent systems. He is author or co-author of more than 20 publications. He is a member of the IEEE Computer Society and the ACM. His current research interests are related to the design and implementation of Distributed Intelligent Systems for e-Manufacturing and e-Learning, and the Agent-based Software Engineering. Douglas H. Norrie currently holds the Nortel Chair in Intelligent Manufacturing at The University of Calgary, Alberta, Canada. Formerly, he was Head of the Department of Mechanical Engineering at The University of Calgary, and more recently Head of the Division of Manufacturing Engineering. He is also Adjunct Professor of Computer Science at the same institution. He has been a member of two major International Research Consortia in Intelligent Manufacturing Systems, being a founder member in the Holonic Manufacturing Systems and the Gnosis Knowledge Systematization consortia. He is author or co-author of numerous publications, including seven books. His research interests are in Intelligent Systems and, in particular, in Multi-Agent Systems.
PY - 2001/10
Y1 - 2001/10
N2 - In agent-oriented manufacturing systems, an efficient and flexible computational conversation model can enable more effective and robust communication, cooperation, and negotiation among agents. This paper therefore describes a schema-based agent conversation model for intelligent agent-oriented manufacturing systems. Conversation schemata are constructed by (1) identifying agent types and their interactions; (2) capturing and formalizing interaction constraints; (3) verifying schemata using colored petri net (CPN) (to prevent deadlock and livelock of conversations); (4) translating the schemata into production rule sets and Java thread classes; (5) identifying inter-schema relationships and building class hierarchies of schemata; (6) constructing conversation managers (CMs) to control and coordinate the schema thread-based dynamic execution. An application example in distributed production planning, implemented in the Java environment, shows the effectiveness of the method.
AB - In agent-oriented manufacturing systems, an efficient and flexible computational conversation model can enable more effective and robust communication, cooperation, and negotiation among agents. This paper therefore describes a schema-based agent conversation model for intelligent agent-oriented manufacturing systems. Conversation schemata are constructed by (1) identifying agent types and their interactions; (2) capturing and formalizing interaction constraints; (3) verifying schemata using colored petri net (CPN) (to prevent deadlock and livelock of conversations); (4) translating the schemata into production rule sets and Java thread classes; (5) identifying inter-schema relationships and building class hierarchies of schemata; (6) constructing conversation managers (CMs) to control and coordinate the schema thread-based dynamic execution. An application example in distributed production planning, implemented in the Java environment, shows the effectiveness of the method.
KW - Colored petri nets
KW - Intelligent agent
KW - Intelligent manufacturing system
KW - Schema
UR - http://www.scopus.com/inward/record.url?scp=0035474354&partnerID=8YFLogxK
U2 - 10.1016/S0166-3615(01)00133-6
DO - 10.1016/S0166-3615(01)00133-6
M3 - Journal Article
AN - SCOPUS:0035474354
SN - 0166-3615
VL - 46
SP - 259
EP - 274
JO - Computers in Industry
JF - Computers in Industry
IS - 3
ER -