Foundation for Intelligent Physical Agents (FIPA)

Foundation for Intelligent Physical Agents (FIPA) Agent advantages:  ‘Natural’ Abstraction level  Easy to specify knowledge  Useful breakdown of problems  Modularity  Coping with legacy systems Why agents are useful ? The agents are useful because intelligent agents...

Foundation for Intelligent Physical Agents (FIPA) Agent advantages:  ‘Natural’ Abstraction level  Easy to specify knowledge  Useful breakdown of problems  Modularity  Coping with legacy systems Why agents are useful ? The agents are useful because intelligent agents make software more adaptable, connected, compatible, and capable of handling complex tasks intelligently. Why agents are not useful ? The agents are not useful because the wizards, scripting languages, programming languages, neural networks, pure rule-based systems and object. The Human Agents: The most important characteristics of the human agent is that human agents are goal-driven, adaptable, interactive and intelligent beings. Agent standards: FIPA, JAS, OMG Agent initiatives: Agenticities, TILAB and MOTOROLA Robust agent platforms: JADE + J2EE -> BlueJADE The Agent has only ONE agent management system. The Agent Platform supports the message transport system and the message transport is the default communication method between agents on different Agent Platforms. Agent Management System (AMS) and Directory Facilitator (DF) have a common subset of actions: register, deregister, search and modify. What is FIPA ? The main goal of FIPA is to define a full set of standards for implementing systems within which agents could execute and its specifying how agents should communicate and interact. FIPA Agent Execution Model The agent completely its thread of execution, and decides itself when to read received messages and which messages to read. The transport mechanism filles a private queue but it does not call the agent code. Agent needs concurrency that can change in multiple simultaneous conversation and can execute several concurrent tasks. Agent Naming Agent is identified through an extensible collection of parameter value pairs, called an Agent Identifier (AID). The name of an agent is immutable: the other parameters in the AID of an agent can be changed. A given agent may support many methods of communication and can put multiple transport address values in the addresses parameter of an AID. We have JADE states like ap_initiated, ap_active, ap_suspended, ap_waiting, ap_deleted, ap_transit. We have two more ap_copy and ap_gone but they are generated from the JADE platform. Agent Communications – Messaging Message transport protocols: IIOP, HTTP Three encoding: String encoding, bit-efficient encoding, XML. Each message protocol supports the three kinds encoding. Message Structure: A message is made up of two parts: 1. Envelope that is used to transport information 2. Body is used for comprising the ACL message of the agent communaction Agent Communication Language – ACL: We use ACL for structured message, performative, addressing etc. and we have so many options we can just in ACL that is: in reply to, reply by, in reply to, content, reply with. (its like when we want to send an email.) Content Language – CL: Defines: proper syntax for content expressions, number of operators such as logical connectors AND and OR. Semantic Language – SL: Describes actions, propositions and identifying expressions