430 likes | 562 Vues
Adaptive Intelligent agent in real-time strategy games. An Introduction. Project Members. Omar Enayet. Abdelrahman Al- Ogail. Ahmed Atta. Amr Saqr. Dr. Mostafa Aref. Dr. Ibrahim Fathy. Agenda. Problem Definition Theoretical Areas of Problem. Project Domain.
E N D
Adaptive Intelligent agent in real-time strategy games An Introduction
Project Members Omar Enayet Abdelrahman Al-Ogail Ahmed Atta AmrSaqr Dr. MostafaAref Dr. Ibrahim Fathy
Agenda • Problem Definition • Theoretical Areas of Problem. • Project Domain. • Specific Problem Definition. • Motivations. • Objectives. • Project Background • Survey & Approach. • Domain Platform. • AI Engine Architecture. • Expected Deliverables. • Development Tools. • Project Time Plan. • Web Resources & References.
Theoretical Areas Learning Make the machine learn.
Theoretical Areas Planning Plan then re-plan according to new givens.
Theoretical Areas Knowledge Sharing Let everyone know instantly what you knew through experience.
Theoretical Areas - Summary • Learning • Planning • Knowledge Sharing
Project Domain RTS Games Real-Time Strategy Games.
Problem Definition Predictability Computer Opponent actions easily predicted.
Problem Definition Non-Adaptability Computer Opponent doesn’t adapt to changes in human actions.
Problem Definition Static Scripts Computer AI relies on static scripting techniques.
Problem Definition Experience Loss The Absence of sharing experience costs a lot.
Problem Definition - Summary • Predictability • Non-Adaptability • Static Scripting • Experience Loss
Motivations Interested In Machine Learning
Motivations Interested In RTS Games
Motivations Meets Our Career Ambitions as AI Programmers
Why this Domain Rich Environment. Severe Time Constraints – Real-Time AI – Many Objects – Imperfect Information – Micro-Actions
Why this Domain Active Research Theses and Papers are from 2003 to 2009.
Why this Domain Wars Simulation Research in this field contributes to the modern warfare Research. Half of The USA’s Army will be robots in the coming years.
Objectives Adaptive A.I. Making the Computer Opponent adapt to changes like human do.
Objectives Mobile Experience Making Sharing Experience Possible Among Machines
Survey • Adaptive Reinforcement Learning Agents in RTS Games Recent Papers • Case-based planning and execution for real-time strategy games. • Transfer Learning in Real-Time Strategy Games Using Hybrid CBR/RL We Collected more than 30 papers concerning this field, different in their way of approaching the problem and the techniques used to solve the problem. Examples of them are above ^
Survey Asking Experts Alex Champandard Eric Kok Alex Champandard, Eric Kok and more.
Survey Reference Book The Book “AI Game Engine Programming” talks about the drawbacks in learning and planning in RTS Games.
Approaches - Techniques Reinforcement Learning A Sub-Science of Machine Learning.
Approaches - Techniques Case-Based Planning Planning using Case-based reasoning.
Approaches - Techniques BDI Agents Tech. Beliefs-Desires-Intentions Agents.
Techniques - Summary • Reinforcement Learning • Case-Based Planning • BDI-Agent Tech.
Approaches - Languages C++ The Main Language our Open Source Game is coded with.
Approaches - Languages 2APL An Agent-Oriented Language.
Approaches - Languages LUA A Scripting Language widely used in Video-Games.
Platform – Not Chosen ORTS An Open-Source RTS Game
Platform – Not Chosen Wargus An Open-Source RTS Game based in Stratagus Game Engine
Platform – The Chosen One BosWars An Open-Source RTS Game based in Stratagus Game Engine
Game Engine AI Engine That’s our guy
Expected Deliverables Enhanced AI Engine An AI Engine which makes the computer behavior in the game as human as possible.
Expected Deliverables Experimental Results Comparison of the results of the enhanced AI Engine with ordinary static AI.
Development Tools • Visual Studio 2008 Professional Edition • 2APL Environment • C++ Libraries : Boost, Guichan ..etc.
Web Resources • Project Blog : http://rtsairesearch.wordpress.com/ • SVN Repository : https://mzrtaiengine.googlecode.com/svn/trunk/
References • Book : AI Game Engine Programming • Book : Artificial Intelligence for Games • The Most Important Papers/Theses : Thanks • Eric Kok - Adaptive Reinforcement Learning Agents in RTS Games –– Master Thesis – University of Utrecht - 2008 • SantiOnta˜n´on, KinshukMishra, NehaSugandh, and Ashwin Ram. Case-based • planning and execution for real-time strategy games. In Proceedings of ICCBR - 2007 - 2007 • Manu Sharma, Michael Holmes, Juan Carlos Santamaria, AryaIrani, Charles Lee Isbell Jr., Ashwin Ram: Transfer Learning in Real-Time Strategy Games Using Hybrid CBR/RL. IJCAI 2007: 1041-1046
Web References • http://gadgets.softpedia.com/news/US-Soldiers-Will-be-Half-Robots-Half-Human-by-2015-1334-01.html • http://www.france24.com/en/20090205-sciences-usa-robot-future-american-army-videogame-soldiers-machine