Comparing reactive techniques to classical planning for intelligent virtual agents

• Název práce v češtině: Comparing reactive techniques to classical planning for intelligent virtual agents
• Název v anglickém jazyce: Comparing reactive techniques to classical planning for intelligent virtual agents
• Klíčová slova: inteligentní virtuální agenti, specifikace chování, klasické plánování, reaktivní techniky
• Klíčová slova anglicky: intelligent virtual agents, behvaior specification comparism, classical planning, reactive techniques
• Akademický rok vypsání: 2011/2012
• Typ práce: diplomová práce
• Jazyk práce: angličtina
• Ústav: Katedra softwaru a výuky informatiky (32-KSVI)
• Vedoucí / školitel: Mgr. Jakub Gemrot, Ph.D.
• Řešitel: skrytý - zadáno a potvrzeno stud. odd.
• Datum přihlášení: 30.01.2012
• Datum zadání: 13.02.2012
• Datum potvrzení stud. oddělením: 16.02.2012
• Datum a čas obhajoby: 03.09.2012 09:30
• Datum odevzdání elektronické podoby: 03.08.2012
• Datum odevzdání tištěné podoby: 03.08.2012
• Datum proběhlé obhajoby: 03.09.2012
• Oponenti: RNDr. Tomáš Balyo, Ph.D.

Zásady pro vypracování

The scope of the thesis is to compare reactive techniques to classical planning
for the purpose of control of intelligent virtual agents (IVAs) in 3D real-time
dynamic environments.

Reactive techniques are dominant paradigm for creation of IVAs
controllers, however they have their limitations due to their fixed plans, which cannot be
altered during runtime. Therefore IVAs can benefit from exploiting techniques of classical planning.

Unfortunately only few attempts were made to employ classical planning
for controlling IVAs. There is also lack of reliable performance comparison of
planning algorithms and reactive agent architectures.

The goal of the thesis is to connect several existing classical
planners and planners for relaxed domains to Pogamut, a platform
for development of IVAs for virtual environment of Unreal Engine,
and to design a suitable test environment that would allow to compare their
performance in solving tasks of varying difficulty under different
levels of dynamicity. The thesis will analyze the performance of
IVAs employing classical planners to IVAs with reactive
architecture and to IVAs that employ planners for relaxed domains.

Seznam odborné literatury

Barták, R., Dvořák, F., Gemrot, J., Brom, C., Toropila, D.:
When Planning Should Be Easy: On Solving Cumulative Planning Problems,
PlanSIG-2011 (accepted).

Gemrot, J., Kadlec, R., Bida, M., Burkert, O., Pibil, R., Havlicek,
J., Zemcak, L., Simlovic, J., Vansa, R., Stolba, M., Plch, T., Brom
C.: Pogamut 3 Can Assist Developers in Building AI (Not Only) for
Their Videogame Agents. In: Agents for Games and Simulations, LNCS
5920, Springer, 2009, pp. 1-15.

Porteous, J., Teutenberg, J., Charles, F., Cavazza, M.:
Interactive Storytelling with Temporal Planning (Demo Paper). In:
Proceedings of the 10th International Conference on Autonomous Agents
and Multiagent Systems (AAMAS - Demos), Taipei, Taiwan, May 2011.

Jeff Orkin: Three States and a Plan: The A.I. of F.E.A.R., GDC 2006

Chih-Wei Hsu and Benjamin W.Wah: The SGPlan Planning System in IPC-6

Russell, S., Norvig, P.: Artificial Intelligence: A Modern Approach, 1995