Research in Games Technology at the ACRI Games Technology Laboratory

In the ACRI Games Technology Laboratory there is a large range of research projects underway and planned.

1. Open Game Development System
This is a completed pilot project to test the possibility of coherent game development by developers who are spread across the globe. All tools, game documents and assets are centrally located and exposed through a web interface. The tools and software are platform dependent and yet remote developers can run them as if they were running on their local platform even though it may not be a platform compatible with the game tool software being used.

2. GT Tool Extensions
Tools are essential to game development work. The tools need to integrate with the game engine and related software – that is they should create assets that are compatible and work with other software. The separate tools do a single function. This makes them easier to update and they can be used in any order similar to the Unix concept of system tools. Outputs from one tool become inputs to other tools. The way tools can be connected and used and the needs for new tools and the way we use them is an area of study. Tools can always be improved but we must avoid making them so complex that their utility decreases. We have tools for building tiling maps, tools for creating fractal horizons and 3D mountain ranges, basic polygon design tools, animation testing tools, tools for sound file generation by formula and tools for random backdrop creation amongst many others.

3. Plug-in Audio Scripting Language
Game audio will need to become programmable – generated within the game appropriate to the situation, rather than being canned. A full-powered object oriented language with the capabilities of C++ and integrable with C++ will be needed. Research on this topic has been underway for 9 months already.

4. Higher Dimensional Geometric Algorithms using Geometric Algebra
The transition from 2D games to 3D games has been a large step with considerable change in the software libraries and interfaces. 3D brought new algorithmic problems and processing trade-offs for interactive game software. We are looking at the transition to higher dimensions and new programming methodologies beyond quaternions (such as multivectors) that will reduce the complexity of multidimensional geometric algorithms and computations.

5. New Approaches to PBC (Physically-Based Computing)
Computation of game physics has been exceedingly expensive within games but it is needed for greater realism. We are looking at new ways to deal with collision detection and collision response that can reduce the relative demand on CPU time.

6. Next Generation GT Hardware Technology Support
Hardware is continually improving in speed. Graphics cards have been improving in the polyhedral rendering paradigm. However future machines may well employ interactive ray-tracing which would make polygon engines processors obsolete. We have been researching new space partitioning algorithms that will assist future ray-tracing rendering technology. Additionally we have research plans for true 3D hardware rendering systems. In particular we have started research on computer generated holographic animation systems.

7. Game Architectures and Language Structure Usage (Advanced C++ Features)
We have been evaluating the new advanced C++ programming language features for their usefulness in game software designs. Language design, evaluation and implementation have always been corner stones of Computer Science research. Software Engineering is also now an established and mature research area in Computer Science with particular implications for Game Architecture. By applying these research results and principles of Computer Science, we want to ensure that our students use the best of the language for rapid high-end game development and avoid the hasty development of games without a proper or indeed any apparent architecture.

8. Comparative Game Engine Analysis and Design
Game Engines from low-end to high-end are becoming widely available with new versions having significant improvements over the previous versions. In this project we want to compare technical features, ease of use, performance and limitations of games engines. Some engines are more appropriate to low memory systems and some are appropriate to high quality workstations. We have also created our own high-end games engine (Ascent 3D) with real-time bump-mapping and shadow casting. This research concerns the architectural design view of the engine code with a view to modularity and mix-and-match engine construction.

9. Bringing Traditional CG Algorithms into GT
There are many algorithms in traditional Computer Graphics that have not yet been brought into games because the processing time has been prohibitive. We have been doing research on testing and optimizing these algorithms for use in real-time applications.

10. Bringing Traditional AI Algorithms into GT
There are many algorithms in traditional Artificial Intelligence that have not yet been brought into games because the processing time has been prohibitive. Examples of these are Natural Language Processing and Generation, Neural Network simulations, Evolution Strategies and Genetic Algorithms. We have been doing research on modifying these algorithms so that they can be used to good effect in games.

11. Component-based Approach to Game Building
The visual languages such as Visual Basic, Delphi and Borland C++ Builder have shown how software GUI development can be done more rapidly. True visual languages have no code editor interface at all and simply connect icons together. This is a powerful paradigm that can be used for game development too. We are researching the development of Active Game Components that can be placed and connected on a backdrop for fast level design. These components represent game entities including both moving and fixed entities.

12. New Realistic Graphics Effects
Special graphics such as mirages, fire, waterfalls, spray and water surfaces have been already researched and made available for game technologists. This research is never complete however as various approximations of reality are always being improved upon in striving for ever more realistic effects.

Information provided by Dr John R Rankin
Head of the ACRI Games Technology Laboratory
Department of Computer Science and Computer Engineering
La Trobe University
Bundoora, Australia 3086.