Research Issues of RoboCup Physical Agent Track

 

RoboCup physical agent challenges are summarized as follows:

• Perception: The player should observe in real-time the behaviors of other objects which it cannot completely predict or control in order to take an action to deal with them. Such objects include a ball, an opponent, and in some sense, a common-side player, and further a judge. Capabilities of wide range perception, discrimination of other agents, and estimation of their locations and motions coping with occlusions are needed. Such perception is a basic technology to expand robotic applications.
• Actions: Controlling a ball in a multi-agent situation introduces a new set of problems even in robotics, where traditionally the manipulated objects are mostly stationary or follow predictable trajectories. Some of the immediate challenges are discussed in depth in the following sections. Long term challenges including: a) controlling a 3D (bouncing or flying) motion of a ball, b) soccer playing with limbed (quadrupeds, hexapods, or even bipeds) robots, and c) generating "faint" actions assuming that the opponent has a capability of action observation and prediction.
• Situation and Behavior: The task domain itself is simple, but infinitely many situations can occur in accordance with dynamic changes of the relationships in position and relative motion between the ball, the goal, and the players, and the context of the game. The optimal behavior changes from one situation to another. Since our goal is a soccer playing team, we need abilities beyond simple reflexive behaviors. For eample, we need situation understanding, tactics selection and modification, minimum communication with teammates, teamwork behaviors acquired through practical training. These issues are closely related to the cognitive issues such as organization of spatio-temporal memory of the world and categorization of sets of motor behaviors into skills (symbols) grounded by the physical bodies.
• Real-Time: Since the situation rapidly changes according to motions of the ball and other players, there is no time to carefully analyze the situation and deliberate a plan. Therefore, the player should take an action such as kicking a ball immediately or dribbling it if surrounded by no opponents in order to wait for a better situation, or moving to a certain position in order to facilitate future or collaborative actions.
• Platform: The initial barrier for physical agent track challengers would be to build a team of robotic platforms for soccer playing. A platform should have a power source, high speed and quick response mobility, a ball kicking capability, sensors for a ball, the goal, the soccer court, self position, and other players' position and movements, on-board and/or remote computers, and a wireless communication for at least accepting the judge's commands. Integrating all these into a size-constrained physical structure while achieving entire real-time performance is already a challenge which requires constrained optimization of a large number of design parameters. Participating in RoboCup requires producing and maintaining multiple such copies. A valuable short range challenge would be to propose a standard robotic platform design for robot soccer. Although this is not exactly an AI issue, such a standard platform will surely facilitate many AI researchers potentially wanting to test their ideas on real robots.

These challenges described above are significantly long-term ones along the way to realizing a good soccer-playing robot team. Some of them will take a few decades to meet. However, due to the clarity of the final target, several subgoals can be derived, which define mid term and short term challenges. One of the major reasons that RoboCup is attractive to so many researchers is that it requires the integration of a broad range of technologies into a team of complete agents, as opposed to a task-specific functional module. The long term research issues are too broad to compile as a list of specific items. Nevertheless, the challenges involve a broad range of technological issues ranging from the development of physical components, such as high performance batteries and motors, to highly intelligent real-time perception and control software.

The mid-term technical challenges, which are the target for the next 10 years, can be made more concrete, and a partial list of specific topics can be compiled. Following is a partial list of research areas involved in RoboCup physical agent track, mainly targeted for the mid term time span: (1) agent architecture in general, (2) implementation of real-time and robust sensing, (3) realization of stable and high-speed robot control, (4) sensor fusion, (5) behavior learning for multi agent environments, and (6) cooperation in dynamic environments.

The RoboCup Physical Agent Challenge should be understood in the context of larger and longer range challenges, rather than as a one-shot challenge. Thus, we wish to provide a series of short-term challenges, which naturally leads to the accomplishment of the mid- term and long-term challenges.