Software intensive systems are increasingly expected to deal with changing user needs and dynamic operating conditions at run time. Examples are the need for life reconfigurations, management of resource variability, and dealing with particular failure modes. Endowing systems with these kinds of capabilities poses severe challenges to software engineers and necessitates the development of new techniques, practices, and tools that build upon sound engineering principles. The field of multi-agent systems focuses on the foundations and engineering of systems that consists of a network of autonomous entities (agents) that int- act to achieve the system goals. One line of research in multi-agent systems, inspired by biological, physical and other naturally occurring systems, concerns multi-agent systems in which agents share information and coordinate their - havior througha shared medium called an agentenvironment. Typical examples are gradient fields and digital pheromones that guide agents in their local c- text and as such facilitate the coordination of a community of agents. Since environment-mediation in multi-agent systems has shown to result in mana- able solutions with very adaptable qualities, it is a promising paradigm to deal with the increasing complexity and dynamism of distributed applications. Control in environment-mediated multi-agent systems is decentralized, i. e. , noneof the components has full accessor control over the system. Self-organization is an approach to engineer decentralized, distributed and resource-limitedsystems that are capable of dynamically adapting hanging conditions and requirements without external intervention. This useful system property is often reffected in functionssuchasself-configuration,self-optimization,andself-healing. Engine- ing approachesto self-organizing systems often rely on global functionality to emerge from localand autonomous decisions of individual agents that commu- catethroughasharedagentenvironment.