Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.498345
Title: Understanding multi-agent design as coordination
Author: Alexiou, Aikaterini
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Design decision making increasingly involves the participation of multiple agents which bring into the design process multiple, and often conflicting, needs, knowledge, and goals. To the human agents (experts from the same or different domains, clients, users, stakeholders) one should add artificial agents (computational models and tools more generally) that play an important part in the process. Design research has considered the issue of distributed decision making mainly through the concepts of cooperation and collaboration. The present thesis argues that coordination is a more apposite concept for capturing the social distributed character of design. The concept of coordination places emphasis on issues of interdependency, complexity and distribution and enables us to understand design at a systemic/organisational level, without making assumptions about agents' commitment to a common goal, or their disposition towards cooperation or conflict. Additionally, coordination is used to capture the generative, creative aspects of distributed design decision making. The study explores and establishes the meaning of coordination through experimentation with computational models and simulations. The very process of building these models is a vehicle for exploring key hypotheses and assumptions, and developing a coherent theoretical construction. Overall, the thesis identifies the key dimensions of coordination that are typical to the domain of design, and employs them to develop a framework (a theory) for understanding multi-agent design as a generative social process. The dimensions identified are learning, decentralised control and co-evolution. A model of coordination developed using the paradigm of distributed learning control is used as a way to establish the precise meaning of these dimensions. Based on insights from the experimentation, the concept of coordination is further refined in order to propose an organisational (complexity-informed) perspective of multi-agent design. According to this perspective, the relationship between agents, their goals, and the design variables they manipulate, is at the same time a product of the design process, but also a constraint over individual agents. Coordination is then defined as a dynamic process towards a scheme of organisation that entails the emergence of collective design solutions.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.498345  DOI: Not available
Share: