Multi-Agent Systems, and Holonic Multi-Agent Systems, mainly exploited and developed in Manufacturing Systems, are more and more used in other fields such as complex system simulations, information systems, distributed problems solving...

Holonic Multi-Agent Systems are flexible; their only strict obligation is to follow a pyramidal structure. They really have an interesting potential, however, methods and tools in order to facilitate their design and management, especially regarding to their Self-Adaptation and Self-Organization capabilities, are required. Indeed, the main characteristics of a Holonic Multi-Agent System, that are the robustness and the cooperation, are dependent of the dynamic of the Multi-Agent System and of unexpected events that occur in the environment.

Robustness enables the Multi-Agent Systems to cope with perturbations, faults and hard constraints. Such a property can typically be implemented using self-organization solutions such as regeneration of faulty agents, generation of new agents and/or migration of agents between Multi-Agent Systems.

Cooperation ensures that if each agent, classically pursuing its own goals and strategies in the pyramidal organization, interacts with each others in a cooperative way, all agents will collectively lead the Multi-Agent System to its intended overall goal. This property is generally processed by locally and autonomously modifying agents' roles or behaviours, by adapting their strategies; it is a way to implement self-adaptation of the agents/holons.

This Special Session of SOHOMA welcomes models and methods linked to robustness, cooperation, self-adaptation and self-organization in Multi-Agent Systems, enabling emergence of new behaviours as well as sub-organizations.

---- Publications
All papers duly presented at the Workshop will be included in the :
SOHOMA'14 post conference Proceedings publication in special issues of: Springer series "Studies in Computational Intelligence", Journal of Intelligent Manufacturing (Springer) and Computers in Industry (Elsevier).
----