NETWORKS OF AGENTS

Our society is like a brain, made up of massive networks of people, organizations, supply chains, local communities, companies, etc. Like the neurons in our brain, people and organizations are constantly forming new connections and interact with other people and organizations. However, left to themselves, people cannot achieve the same rates and frequencies of communication that the neurons in our brain can.

Quick communication

If we allow objects, people, organizations and their goals to be represented by software agents, they can accomplish much more. Agents can communicate in 'hypertime', and thus communicate and negotiate much quicker than humans.

Agents interact with people to achieve specific goals. Performance becomes even better in a network of agents, where software agents can negotiate with other agents, instead of having to wait for an actual person to respond.

Collective brain

This way, agent networks help people and their communities in forming a collective brain. If people in a professional or social network share access to a common knowledge-base, they can cooperate much more effectively. Agents can filter information, and make sure that people receive only relevant messages, and do not succumb to an overload of “noise” or useless information.

Agent networks -or multi-agent systems- can thus be beneficial in many professional domains: from incident management to logistics and from health care to the workforce deployment.

Challenges:

> How do we allow agents to learn from each other?

> How do we group people and their personal agents, so these agents may use feedback from others?

> How do we interpret and filter information?

Keywords:

> multi-agent systems

> fast and frequent communication

> negotiation

> collective knowledge