Jakub Gemrot

Pogamut 3 Can Assist Developers in Building AI (Not Only) for Their Videogame Agents

Many research projects oriented on control mechanisms of virtual agents in videogames have emerged in recent years. However, this boost has not been accompanied with the emergence of toolkits supporting development of these projects, slowing down the progress in the field. Here, we present Pogamut 3, an open source platform for rapid development of behaviour for virtual agents embodied in a 3D environment of the Unreal Tournament 2004 videogame. Pogamut 3 is designed to support research as well as educational projects. The paper also briefly touches extensions of Pogamut 3; the ACT-R integration, the emotional model ALMA integration, support for control of avatars at the level of gestures, and a toolkit for developing educational scenarios concerning orientation in urban areas. These extensions make Pogamut 3 applicable beyond the domain of computer games.

A periphery of Pogamut: from bots to agents and back again

Despite virtual characters from 3D videogames - also called bots - seem to be close relatives of intelligent software agents, the mechanisms of agent reasoning are only rarely applied in videogames. Why is this? One possible reason is the incompatibility between representations used by agent decision making systems (DMS) and videogame worlds, as well as different handling of these representations. In recent years, we developed Pogamut, which is a toolkit for coupling videogame worlds with DMSs originating within the agent oriented research as well as other disciplines, allowing for controlling in-game characters by these DMSs. To this end, Pogamut features an interface bi-directionally bridging the representational gap between a game world and an external DMS. This paper conceptualises functionality of this interface based on our experience with connecting Pogamut to various game worlds, most notably Unreal Tournament 2004. We present a general abstract framework, which verbalises requirements an agent researcher must fulfil in order to employ his/her reasoning mechanism for controlling in-game virtual characters. This paper also reviews Pogamut, which the researcher can utilise.

Towards Fast Prototyping of IVAs Behavior: Pogamut 2

We present the platform for IVAs development in the human like environment of the first-person shooter game Unreal Tournament 2004. This environment is extendible and supported by vast community of users. Based on our previous experience the problem of fast verification of models of artificial intelligence or IVAs is in implementation issues. The developer spends most of his time solving technical environment dependent issues and malfunctions, which drives him away from his goals. Therefore our modular platform provides a tool, which helps solving those problems and the developer can spend saved time by solving another AI based issues and model verification. The platform is aimed for research and educational purposes.

How to compare usability of techniques for the specification of virtual agents' behavior? an experimental pilot study with human subjects

Reactive or dynamic planning is currently the dominant paradigm for controlling virtual agents in 3D videogames. Various reactive planning techniques are employed in the videogame industry while many reactive planning systems and languages are being developed in the academia. Claims about benefits of different approaches are supported by the experience of videogame programmers and the arguments of researchers, but rigorous empirical data corroborating alleged advantages of different methods are lacking. Here, we present results of a pilot study in which we compare the usability of an academic technique designed for programming intelligent agents behavior with the usability of an unaltered classical programming language. Our study seeks to replicate the situation of professional game programmers considering using an unfamiliar academic system for programming in-game agents. We engaged 30 computer science students attending a university course on virtual agents in two programming assignments. For each, the students had to code high-level behavior of a 3D virtual agent solving a game-like task in the Unreal Tournament 2004 environment. Each student had to use Java for one task and the POSH reactive planner with a graphical editor for the other. We collected quantitative and qualitative usability data. The results indicate that POSH outperforms Java in terms of usability for one of the assigned tasks but not the other. This implies that the suitability of an AI systems-engineering approach is task sensitive. We also discuss lessons learnt about the evaluation process itself, proposing possible improvements in the experimental design. We conclude that comparative studies are a useful method for analyzing benefits of different approaches to controlling virtual agents.

Emohawk: Searching for a Good Emergent Narrative

We report on the progress we have achieved in development of Emohawk, a 3D virtual reality application with an emergent narrative for teaching high-school students and undergraduates the basics of virtual characters control, emotion modelling, and narrative generation. Besides, we present a new methodology, used in Emohawk, for purposeful authoring of emergent narratives of Facades complexity. The methodology is based on massive automatic search for stories that are appealing to the audience whilst forbidding the unappealing ones during the design phase.

Stories from the history of czechoslovakia , a serious game for teaching history of the Czech lands in the 20th century: notes on design concepts and design process

In the context of curricular history education both commercial entertainment games as well as serious games specifically tailored for educational purposes were employed. Especially the latter types of games were reported as being promising concerning instructional effectiveness. Still, there are not many complex serious games for history education, particularly in the secondary schools context. In this work-in-progress paper, we report on the progress of project Stories from the History of Czechoslovakia, a serious game for teaching history of the Czech lands in the 20th century. We introduce main game concepts, describe two main design challenges we have been facing during the development and how we have addressed them and overview our feasibility study on 71 high-school students. This paper can be informative for researchers and designers working on similar projects.

Notes on pragmatic agent-programming with Jason

AgentSpeak(L), together with its implementation Jason, is one of the most influential agent-oriented programming languages. Besides having a strong conceptual influence on the niche of BDI-inspired agent programming systems, Jason also serves as one of the primary tools for education of and experimentation with agent-oriented programming. Despite its popularity in the community, relatively little is reported on its practical applications and pragmatic experiences with adoption of the language for non-trivial applications. n nIn this paper, we present our experiences gathered during an experiment aimed at development of a non-trivial case-study agent application by a novice Jason programmer. In our experiment, we tried to use the programming language as is, with as few customisations of the Jason interpreter as possible. Besides providing a structured feedback on the most problematic issues faced while learning to program in Jason, we informally propose a set of ideas for solving the encountered design problems and programming language issues.

When a couple goes together: walk along steering

Steering techniques are widely used for navigation of single agents or crowds and flocks. Steerings also have the potential to coordinate movement of human-like agents in very small groups so that the resulting behavior appears socially believable, but this dimension is less explored. Here, we present one such “social” steering, the Walk Along steering for navigating a couple of agents to reach a certain place together. The results of a believability study with 26 human subjects who compared the new steering to the known Leader Following steering in eight different scenarios suggest the superiority of the Walk Along steering in social situations.

3D Immersion in Virtual Agents Education

Many projects featuring intelligent virtual agents have emerged in last years, but not many reports on their advances in education. This paper presents the curricula of a university course on Modelling Behaviour of Human and Animal-like Agents, including a seminar in which students develop their own virtual agents using a toolkit we have developed. This course has been also scaled for a workshop with computer science high-school students. An evaluation of the course is presented and main lessons learned overviewed. The paper also explicitly formulates the teaching methodology underpinning the course and outlines several general questions hoping to start a broader discussion on educational issues related to the field of intelligent virtual agents.

Does high-level behavior specification tool make production of virtual agent behaviors better?

Part of Academias motivation for improving agent-based languages and architectures is the hope that some of its solutions might be, one day, adopted by the videogame industry for specification of agents behavior. However, the games industry already employs its own techniques for that purpose. Thus we need rigorous empirical data to provide an insight to the expected utility of academic solutions and pinpoint their most promising features. Game pro-grammers often face situations when they have to understand and modify the work of current or former colleagues, or to extend their own work from months or even years ago. Thus, one way an academias solution could provide value would be that it outperforms the industrys typical approach under these circumstances --- offering better code readability and maintainability. Here we present results of an empirical study modeling this problem. We adopt Java programming as the industry approach (modeling scripting) and choose the POSH reactive planner as an academic approach. We engaged 22 computer science students attending a university course on virtual agents on two programming assignments, in which they had to produce specific high-level behaviors of 3D virtual agents solving different game-like tasks. First, students had to produce the behavior of a particular task from scratch. Second, these behaviors were used in an assignment where students had to extend the behavior coded by someone else. Finally, three months later, 8 of these students were told to extend their own behavior they coded in the first assignment. The quantitative results indicate that Java is as good as POSH in terms of subjective programmers preference and that there is no objective difference between qualities of created behaviors. The qualitative results suggest several useful but also troublesome features of POSH, some of which are shared by other languages, suggesting possible improvements.