Cancelled
AAAI Workshop on Physics-Based Simulation Games (PBSG-15)

January 25-26, 2015, Austin, Texas, USA
Held in Conjunction with the Twenty-Ninth AAAI Conference on Artificial Intelligence
(AAAI 2015)

Deadline extended to November 1st!

Physics simulation games such as Angry Birds, Cut the Rope, Gears, or Feed Me Oil have become increasingly popular in recent years with the wide availability of touch-screen devices. These games are based on a physics simulator that has complete information about all the physical properties of all game objects and the game world. As such, each move and its consequences can be exactly simulated and displayed, which makes the physical behaviour of the game appear very realistic. These games are easy to play as the possible moves are simple. What makes these games hard and engaging is that the number of possible moves can be very large and effectively infinite, and that the consequences of moves are unknown in advance. The large number of moves is due to the effect of the exact location and/or timing of moves, where small changes may result in differences in the outcome of the physics simulation. Without actually simulating a move its outcome is very hard to predict. This gets even harder if the exact physical properties are unknown in advance and have to be observed and learned, or if the game can only be observed using a vision system – both of which correspond to how humans are playing these games. Playing these games as well as or better than the best human players requires a successful AI agent to solve a number of challenging problems from different areas of AI. The impact of successful solutions to these problems goes way beyond games and will be an essential feature of intelligent agents that interact with the physical world. Several AI competitions, such as Angry Birds (http://www.aibirds.org), Computational Pool (http://www.stanford.edu/group/billiards), Geometry Friends (http://gaips.inesc-id.pt/geometryfriends/), and the Physical Travelling Salesman Problem (http://www.ptsp-game.net/), have been initiated to foster research on this important area.

Topics

We invite high quality work on any aspect of physical simulation games research. Topics include but are not limited to:

detect and identify relevant game objects using vision, including previously unknown objects
learn physical properties of relevant game objects and their effects
learn the physics of a game
infer and approximate the outcome of moves
plan the best move or sequence of moves in a given situation
approximation strategies and dealing with uncertainty
combining simulation and reasoning
methods for experimentally evaluating the performance of agents
system description and evaluation of successful AI agents
analyzing the complexity of game levels or designing provably hard levels
physics simulation games for education or serious games

Format

The one-day workshop will include full and short paper presentations, invited talks, demos, posters and discussions.

Submissions

Papers must be formatted in AAAI two-column, camera-ready style; see the AAAI 2015 Author Kit at http://www.aaai.org/Publications/Templates/AuthorKit.zip for details. Submitted papers may be no longer than 6 pages including references. Submissions should be made through EasyChair. Authors of selected papers will be invited to submit an extended version of their paper to the Special Issue on "Physics-based Simulation Games" of the IEEE Transactions on Computational Intelligence and AI in Games. For further details on the special issue see: http://aibirds.org/PBSG-CfP.pdf.