Reactive/Anytime Agents — Towards Intelligent Agents with Real-Time Performance

نویسندگان

M. Adelantado

S. De Givry

Edouard Belin

چکیده

To face the requirements of complex and time-critical applications, several systems for reasoning about time and in time, have been proposed by the AI community: reactive systems provide reactivity to their environment at the cost of short-sighted decision making, and conversely, deliberative systems involve more sophisticated decision making at the cost of response times. Recently, several AI researchers have proposed the concept of anytime algorithms as a means to meet timing constraints in time-critical applications. An anytime algorithm is an iterative refinement algorithm that can be interrupted and asked to provide an answer at any time. A major requirement of real-time embedded systems that has not been directly addressed by the AI community, is predictability because random timing failures may lead to catastrophic situations. Several reactive languages have been developed to face such real-time requirements for pure reactive agents. The objective of this paper is twofold: first of all, we describe a unified view of anytime computing as a framework for designing deliberative time-constrained agents. Secondly, we show how reactive programming and anytime computing gives the opportunity to associate reactive and deliberative agents to achieve flexibility and responsiveness. More generally, our work is an attempt to motivate a closer relationship between the AI community and the real-time programming community. This work is supported by the French Ministry of Defense (DRET).

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Reactive Teaming for Intelligent Game Characters

Reactive methods for controlling intelligent agents have proven to be a very strong alternative to planning methods, especially in areas such as robotics and game agents, where real-time performance is critical. Despite this success, coordinating teams of reactive agents can be a nontrivial problem. In this work, we present our initial development of a novel method for coordinating reactive age...

متن کامل

Saturne: a Reactive -anytime Programming Model for Intelligent Embedded Real-time Systems 3rd Ieee Workshop on Parallel and Distributed Real-time Systems Saturne: a Reactive -anytime Programming Model for Intelligent Embedded Real-time Systems

متن کامل

RT-MOVICAB-IDS: Addressing real-time intrusion detection

This study presents a novel Hybrid Intelligent Intrusion Detection System (IDS) known as RT-MOVICAB-IDS that incorporates temporal control. One of its main goals is to facilitate real-time Intrusion Detection, as accurate and swift responses are crucial in this field, especially if automatic abortion mechanisms are running. The formulation of this hybrid IDS combines Artificial Neural Networks ...

متن کامل

Real Time Goal Orientated Behaviour for Computer Game Agents

To increase the depth and appeal of computer games, the intelligence of the characters they contain needs to be increased. These characters should be played by intelligent agents that are aware of how goals can be achieved and reasoned about. Existing AI methods struggle in the computer game domain because of the real-time response required from the algorithms and restrictive processor availabi...

متن کامل