Home   

MULTI AGENT SYSTEMS

Multi Agent Systems are were Genetically programmed agents have communication capabilities and they learn and teach each other. These agents are capable of evolution and communication together in a group. They observe, learn and teach each other and these multi agents systems are extremely useful tools of the future. Evolutionary capability makes agents self learning and evolving after learning about their mistakes. In a group they learn from other agents mistakes and evolve. Multi agent systems can be in your computers or in Robots. 

Task allocation and conflict resolution are both key areas in multi agent systems. Genetic Programming can be used to evolve agents that automatically allocate tasks among themselves both statically and dynamically, and can resolve conflicts arising from their interaction. The control strategies included local control, distributed control and central control. In a man machine conflict machines may over rule man at a future date using multi agent systems. If Evolving Defined Functions library is used in multi agent systems then agents are capable of evolving in time learning from their past mistakes and create more fit populations. Imagine a situation only hypothetical that there is real threat to our earth or solar system due to Gravitational imbalances in the Galaxy. Man may built machines with very high capabilities and arrange transportation of human to another planet or Galaxy along with robots to help in hazardous environment. In such a scenario Multi Agent Systems with Communication Channels and EDF may result in Robots either extremely useful or over ruling humans.

Assembly lines for product manufacturing using Robots may find Multi Agent Systems extremely useful. Defense and Security applications cannot be discussed here.