RoboCup 1997 Publications

From RoboCup Wiki
Jump to: navigation, search
Author Title Year Journal/Proceedings Reftype DOI/URL
Andou, T.

Refinement of Soccer Agents' Positions Using Reinforcement Learning

1998 RoboCup-97: Robot Soccer World Cup I, pp. 373-388 inproceedings


Abstract: This paper describes the structure of the RoboCup team, Andhill, which won the second prize in the RoboCup97 tournament, and the results of the reinforcement learning in which an agent receives a reward when it kicks the ball. In multi-agent reinforcement learning, the trade-off between exploration and exploitation is a serious problem. This research uses observational reinforcement to ease the exploration problem.
André, E., Herzog, G. & Rist, T.

Generating Multimedia Presentations for RoboCup Soccer Games

1998 RoboCup-97: Robot Soccer World Cup I, pp. 200-215 inproceedings


Abstract: The automated generation of multimedia reports for time-varying scenes on the basis of visual data constitutes a challenging research goal with a high potential for many interesting applications. In this paper, we report on our work towards an automatic commentator system for RoboCup, the Robot World-Cup Soccer. Rocco (RoboCup-Commentator) is a prototype system that has emerged from our previous work on high-level scene analysis and intelligent multimedia generation. Based on a general conception for multimedia reporting systems, we describe the initial Rocco version which is intended to generate TV-style live reports for matches of the simulator league.
Asada, M., Stone, P., Kitano, H., Drogoul, A., Duhaut, D., Veloso, M., Asamas, H. & Suzuki, S.

The RoboCup Physical Agent Challenge: Goals and Protocols for Phase I

1998 RoboCup-97: Robot Soccer World Cup I, pp. 42-61 inproceedings


Abstract: Traditional AI research has not given due attention to the important role that physical bodies play for agents as their interactions produce complex emergent behaviors to achieve goals in the dynamic real world. The RoboCup Physical Agent Challenge provides a good test-bed for studying how physical bodies play a significant role in realizing intelligent behaviors using the RoboCup framework [Kitano, et al., 95]. In order for the robots to play a soccer game reasonably well, a wide range of technologies needs to be integrated and a number of technical breakthroughs must be made. In this paper, we present three challenging tasks as the RoboCup Physical Agent Challenge Phase I: (1) moving the ball to the specified area (shooting, passing, and dribbling) with no, stationary, or moving obstacles, (2) catching the ball from an opponent or a teammate (receiving, goal-keeping, and intercepting), and (3) passing the ball between two players. The first two are concerned with single agent skills while the third one is related to a simple cooperative behavior. Motivation for these challenges and evaluation methodology are given.
Balch, T.


1998 RoboCup-97: Robot Soccer World Cup I, pp. 181-187 inproceedings


Abstract: Hardware-only development of complex robot behavior is often slow because robot experiments are time-consuming and robotic hardware is subject to failure. The problem is exacerbated in multirobot tasks like robot soccer where successful experiments require several robots to operate correctly simultaneously. Robotics research proceeds more quickly through a cycle of experimentation in simulation and implementation on hardware than through hardware-only implementation. To facilitate research in the RoboCup soccer domain we introduce JavaSoccer, a portable system supporting development of multi-robot control systems in simulation and, potentially, on soccer robots. In simulation, JavaSoccer emulates the dynamics and dimensions of a regulation RoboCup small size robot league game.
Balch, T.

Integrating Learning with Motor Schema-Based Control for a Robot Soccer Team

1998 RoboCup-97: Robot Soccer World Cup I, pp. 483-491 inproceedings


Abstract: This paper describes a reinforcement learning-based strategy developed for Robocup simulator league competition. In overview: each agent is provided a common set of skills (motor schema-based behavioral assemblages) from which it builds a task-achieving strategy using reinforcement learning. The agents learn individually to activate particular behavioral assemblages given their current situation and a reward signal. The entire team is jointly rewarded or penalized when they score or are scored against (global reinforcement). This approach provides for diversity in individual behavior.
Burkhard, H.-D., Hannebauer, M. & Wendler, J.

AT Humboldt — Development, practice and theory

1998 RoboCup-97: Robot Soccer World Cup I, pp. 357-372 inproceedings


Abstract: This article covers three basics of our virtual soccer team AT Humboldt: We describe our development process in the frame of a practical exercise for students. The resulting efficient agent-oriented realization is explained, and we give a theoretical embedding of our planning component based on BDI.
Many ideas and first of all hard implementation work came from Pascal Müller-Gugenberger, Amin Coja-Oghlan, Adrianna Foremniak, Derrick Hepp, Heike Müller and Kay Schröter in a practical exercise. We wish to thank them for their great efforts.
Ch'ng, S. & Padgham, L.

Team Description: Building Teams Using Roles, Responsibilities, and Strategies

1998 RoboCup-97: Robot Soccer World Cup I, pp. 458-466 inproceedings


Abstract: The Royal Melbourne Knights are designed to interact as a team of soccer playing agents. This paper introduces a framework for modelling agents using the concepts of roles, responsibilities and strategies in its control of the agent's motivation, attention and behaviour, respectively. Through the use of strategies; an agent may form teams that coordinate their actions. The high performance team of agents designed for the Royal Melbourne Knights will compete at the real-world soccer simulation tournament, RoboCup'97.
Cheng, G. & Zelinsky, A.

Real-Time Vision Processing for a Soccer Playing Mobile Robot

1998 RoboCup-97: Robot Soccer World Cup I, pp. 144-155 inproceedings


Abstract: This paper describes vision-based behaviours for an Autonomous Mobile Robot. These behaviours form a set of primitives that are used in the development of basic soccer skills. These skills will eventually be utilised in an effect to tackle the challenge that has been put forward by the ldquoThe Robot World Cuprdquo initiative. The focus of the discussion will be on the vision processing associated with these behaviours. Experimental results and analysis of the visual processing techniques are also presented.
Coradeschi, S. & Karlsson, L.

A Role-Based Decision-Mechanism for Teams of Reactive and Coordinating Agents

1998 RoboCup-97: Robot Soccer World Cup I, pp. 112-122 inproceedings


Abstract: In this paper we present a system for developing autonomous agents that combine reactivity to an uncertain and rapidly changing environment with commitment to prespecified tactics involving coordination and team work. The concept of roles in a team is central to the paper. Agents act and coordinate with other agents depending on their roles. The decision-mechanism of an agent is specified in terms of prioritized rules organized in a decision tree. In the decision-tree both reactivity to changes in the environment and commitment to long term courses of actions (behaviors) are present. The coordination between agents is mainly obtained through common tactics, strategies, and observations of actions of team members, rather than explicit communication. Coordinated behaviors can be specified in terms of roles and are encoded simultaneously for all participating agents.
Frank, I.

Football in Recent Times: What We Can Learn from the Newspapers

1998 RoboCup-97: Robot Soccer World Cup I, pp. 216-230 inproceedings


Abstract: This paper uses a basic statistical analysis of newspaper articles to gain an insight into the game of football. Basic features of the game axe established and examined in a way that should give insights to the designers of RoboCup teams and also suggest future developments for the regulations that determine the RoboCup environment. As concrete examples of possible Soccer Server modifications, we suggest touchline coaching, a revised model of stamina, and the inclusion of substitutions.
Fry, J., Huang, L. & Peters, S.

Team Sicily

1998 RoboCup-97: Robot Soccer World Cup I, pp. 450-457 inproceedings


Abstract: Team Sicily, our entry in the RoboCup-97 simulator track, is designed to achieve cooperative soccer behavior using a realistic and efficient balance of communication and autonomy. The team is implemented in the multithreading logic programming language Gaea, and builds on the dynamic subsumptive architecture model developed by Noda and Nakashima [3].
Fujita, M., Kitano, H. & Kageyama, K.

A Legged Robot for RoboCup Based on "OPENR

1998 RoboCup-97: Robot Soccer World Cup I, pp. 168-180 inproceedings


Abstract: We propose a physical robot platform for RoboCup based on OPENR, which is proposed as a standard architecture for Robot Entertainment, a new entertainment field devoted to autonomous robots. The platform will be provided as a complete autonomous agent so that researchers are able to start working within their research fields. It is possible to rewrite all of, or parts of. the software of the RoboCup player. The configuration of the platform is quadruped type with a color camera, a stereo microphone, a loudspeaker, and a tilt sensor. In order to reduce the computational load, we propose that, the walls of the field and the player itself are painted in different colors so that it is easy to identify the position of essential objects.
Igarashi, H., Kosue, S., Miyahara, M. & Urnaba, T.

Individual tactical play and action decision based on a short-term goal — team descriptions of team Miya and team Niken

1998 RoboCup-97: Robot Soccer World Cup I, pp. 420-427 inproceedings


Abstract: In this paper we present descriptions of our two teams that participated in the simulator league of RoboCup 97. One of the teams is characterized by soccer agents that make individual tactical plays without communicating with each other. The other team is characterized by the use of an action-decision algorithm based on a short-term goal and current information. The two teams were among the best of 8 and 16 teams at the competition.
Inden, T. & Takahashi, T.

Team: Kasuga-bitos with modulation of playing

1998 RoboCup-97: Robot Soccer World Cup I, pp. 443-449 inproceedings


Abstract: Two points are paid attention in implementing our team Kasuga-bitos. One is an agent changes its playing style as a game proceeds. The other is that agents cooperate each other using only visual information without any programmed protocols. Their effects are estimated by games with Ogaltes, the champion team in PreRoboCup 96.
Iso, R. & Inazumi, H.

A Multi-Layered Planning Architecture for Soccer Agent

1998 RoboCup-97: Robot Soccer World Cup I, pp. 513-518 inproceedings


Abstract: Based on manual simulation experiments, we propose a type of agent as a MultiLayerd Planning(MLP) Architecture to identify the game situation and create action policy. As a team, we will arrange the various type of agent according to game strategy, and realize semi-cooperative Multiagent model with minimum amount of communication.
Ito, N., Hotta, T. & Ishii, N.

Describing Soccer Game in EAMMO

1998 RoboCup-97: Robot Soccer World Cup I, pp. 492-499 inproceedings


Abstract: Object-oriented techniques are useful in many fields such as database systems, operating systems, programming, etc.. However, the conventional object-oriented models meet difficult to represent autonomy and flexibility of agents. To solve these problems, we proposed ldquoan Environmental Agent Model for Multi-Objectsrdquo(EAMMO). EAMMO consists of three types of agents as follows: (1) an upper-agent describes autonomous and dynamic objects, (2) a lower-agent describes functional and static objects, and (3) an environmental-agent describes the environment including agents. It has a great influence on the agents in the environment. By using our agent model, we successfully describe modeling a soccer game for RoboCup-97. As the result, we confirmed that our agent model is a suitable paradigm to represent multi-objects.
Itsuki, N.

Team GAMMA: Agent Programming on Gaea

1998 RoboCup-97: Robot Soccer World Cup I, pp. 500-507 inproceedings


Kitano, H., Asada, M., Kuniyoshi, Y., Noda, I., Osawai, E. & Matsubara, H.

RoboCup: A Challenge Problem for AI and Robotics

1998 RoboCup-97: Robot Soccer World Cup I, pp. 1-19 inproceedings


Abstract: RoboCup is an attempt to foster AI and intelligent robotics research by providing a standard problem where wide range of technologies can be integrated and examined. The first RoboCup competition was held at IJCAI-97, Nagoya. In order for a robot team to actually perform a soccer game, various technologies must be incorporated including: design principles of autonomous agents, multi-agent collaboration, strategy acquisition, real-time reasoning, robotics, and sensorfusion. RoboCup is a task for a team of multiple fast-moving robots under a dynamic environment. Although RoboCup's final target is a world cup with real robots, RoboCup offers a software platform for research on the software aspects of RoboCup. This paper describes technical challenges involved in RoboCup, rules, and simulation environment.
Kitano, H., Tambe, M., Stone, P., Veloso, M., Coradeschi, S., Osawa, E., Matsubara, H., Noda, I. & Asada, M.

The RoboCup Synthetic Agent Challenge 97

1998 RoboCup-97: Robot Soccer World Cup I, pp. 62-73 inproceedings


Abstract: RoboCup Challenge offers a set of challenges for intelligent agent researchers using a friendly competition in a dynamic, real-time, multi-agent domain. While RoboCup in general envisions longer range challenges over the next few decades, RoboCup Challenge presents three specific challenges for the next two years: (i) learning of individual agents and teams; (ii) multi-agent team planning and plan-execution in service of teamwork; and (iii) opponent modeling. RoboCup Challenge provides a novel opportunity for machine learning, planning, and multi-agent researchers — it not only supplies a concrete domain to evalute their techniques, but also challenges researchers to evolve these techniques to face key constraints fundamental to this domain: real-time, uncertainty, and teamwork.
Kourogi, M., Kawamoto, Y. & Muraoko, Y.

A Method Applied for Soccer's Behaviors Using Proper Feedback and Feedforward Control

1998 RoboCup-97: Robot Soccer World Cup I, pp. 156-167 inproceedings


Abstract: As a practical way of achieving "Athletic Intelligence(AI)", we have tried to build a robot capable of playing soccer. In this paper, we made a mobile robot that can shoot and dribble a ball. In order to shoot the rolling ball, the robot must predict the future position of the ball so that it can move ahead of the ball. That is, the robot should be controlled by feedforward control rather than feedback control because feedback control does not allow enough time to catch the ball. Therefore, we think that it is important that the robot has internal model with which it can predict the target position. As long as the ball is within the field of view, the robot is under feedforward control. At the final stage of shooting, control is switched to feedback to minimum errors. When dribbling the ball through the flags, the robot must move without touching the flags and also keep the ball in front under feedback control. Since the robot has an internal model, it should follow the target using feedback control. We have checked that proper use of both control improves the two athletic movements so the robot must have an internal model that can predict the future state of the target object.
Luke, S., Hohn, C., Farris, J., Jackson, G. & Hendler, J.

Co-Evolving Soccer Softbot Team Coordination with Genetic Programming

1998 RoboCup-97: Robot Soccer World Cup I, pp. 398-411 inproceedings


Abstract: In this paper we explain how we applied genetic programming to behavior-based team coordination in the RoboCup Soccer Server domain. Genetic programming is a promising new method for automatically generating functions and algorithms through natural selection. In contrast to other learning methods, genetic programming's automatic programming makes it a natural approach for developing algorithmic robot behaviors. The RoboCup Soccer Server was a very challenging domain for genetic programming, but we were pleased with the results. At the end, genetic programming had produced teams of soccer softbots which had learned to cooperate to play a good game of simulator soccer.
Matellán, V., Borrajo, D. & Fernndez, C.

Using ABC2 in the RoboCup domain

1998 RoboCup-97: Robot Soccer World Cup I, pp. 475-482 inproceedings


Abstract: This paper presents an architecture for the control of autonomus agents that allows explicit cooperation among them. The structure of the software agents controlling the robots is based on a general purpose multi-agent architecture based on a two level approach. One level is composed of reactive skills capable of achieving simple actions by their own. The other is based on an agenda used as an opportunistic planning mechanism to compound, activate and coordinate the basic skills. This agenda handles actions both from the internal goals of the robot or from other robots. This paper describes the work already accomplished, as well as the issues arising from the implementation of the architecture and its use in the RoboCup domain.
Matsumoto, A. & Nagai, H.

Decision Making by the Characteristics and the Interaction in Multi-Agent Robotics Soccer

1998 RoboCup-97: Robot Soccer World Cup I, pp. 132-143 inproceedings


Abstract: This paper deals with a decentralized decision-making method on multi-agent robotic system. Soccer game is adopted as an example for its dynamic characteristics. The soccer simulator has been designed and implemented which enables us to test various approaches to the multiagent studies. The modeling of capabilities of a single robot and team strategy is explained at first. Each robotic agent has different characteristics, and the decision making by each robotic agents are done by considering its characteristics and interaction between robotic agents. Several simulation experiments with different combinations of the components and formations have been tested. The result of the simulation experiments was that similar combination of the components resulted in better performance, and the formation where robotic agents are located in the center axis between the goals showed better defense behaviors.
Nakamura, T.

Development of Self-Learning Vision-Based Mobile Robots for Acquiring Soccer Robots Behaviors

1998 RoboCup-97: Robot Soccer World Cup I, pp. 257-276 inproceedings


Abstract: An input generalization problem is one of the most important ones in applying reinforcement learning to real robot tasks. To cope with this problem, we propose a self-partitioning state space algorithm which can make non-uniform quantization of the multidimensional continuous state space. This method recursively splits its continuous state space into some coarse spaces called tentative states. It begins by supposing that such tentative states are regarded as the states for Q-learning. It collects Q values and statistical evidence regarding immediate rewards r and Q values within this tentative state space. When it finds out that a tentative state is relevant by the statistical test on minimum description length criterion, it partitions this coarse space into finer spaces. These procedures can make non-uniform quantization of the state space. Our method can be applied to non-deterministic domain because Q-learning is used to find out the optimal policy for accomplishing the given task. To show that our algorithm has generalization capability, we apply our method to two tasks in which a soccer robot shoots a ball into a goal and prevent a ball from entering a goal. To show the validity of this method, the experimental results for computer simulation and a real robot are shown.
Noda, I., Suzuki, S., Matsubara, H., Asada, M. & Kitano, H.

Overview of RoboCup-97

1998 RoboCup-97: Robot Soccer World Cup I, pp. 20-41 inproceedings


Abstract: RoboCup-97, The First Robot World Cup Soccer Games and Conferences, was held at the Fifteenth International Joint Conference on Artificial Intelligence. There were two leagues: real robot and simulation. 10 teams participated in the real robot league and 29 teams did in the simulation league. The World Champions are CMUnited (CMU, U.S.A.) for the Small-Size league, Dreamteam (Univ. of Southern California, U.S.A.) and Trackies (Osaka Univ., Japan) for the Middle-Size league, and At-Humboldt (Humboldt Univ., Germany) for the Simulation league. The Scientific Challenge Award was given to Sean Luke (Univ. of Maryland, U.S.A.) for his genetic programming based simulation team and the Engineering Challenge Awards was given to UttoriUnited (Utsunomiya Univ., Toyo Univ. and RIKEN, JAPAN) and RMIT (Royal Melbourne Institute of Technology, Australia) for designing novel omni-directional driving mechanisms. RoboCup-98, the Second Robot World Cup Soccer Games and Conferences, will be held at the Third International Conference on Multi-Agent Systems at Paris, France in July 1998.
Ohta, M.

Learning Cooperative Behaviors in RoboCup Agents

1998 RoboCup-97: Robot Soccer World Cup I, pp. 412-419 inproceedings


Abstract: In the RoboCup environment, it is difficult to learn cooperative behaviors, because it includes both real-world problems and multiagent problems. In this paper, we describe the concept and the architecture of our team at the RoboCup'97, and discuss how to make this agent learn cooperative behaviors in the RoboCup environment. We test the effectiveness using a case study of learning pass play in soccer.
Oller, A., Garcia, R., Ramon, J.A., Figueras, A., Esteva, S., de la Rosa, J.L., Humet, J. & Acebo, E.D.

Description of Rogi-Team

1998 RoboCup-97: Robot Soccer World Cup I, pp. 286-294 inproceedings


Abstract: Multi-agent decision-making structures are expected to be extensively applied situations in complex industrial systems (namely distributed systems like energy or material distribution networks, big plants with several stations and difficult processes with huge number of variables). Multi-robotic soccer competitions are a forum where rules and constraints are good to develop, apply, and test those structures. This paper shows new developments under Matlab/Simulink that allow decision-making among agents which, in this case, co-operate to play soccer game.
This work has been funded by the CICYT TAP97-1493-E project "Desarrollo de una segunda generación de equipo microrobótico. Consolidación de la competitividad internacional", of the Spanish government.
Pagello, E., Montesello, F., D'Angelo, A. & Ferrari, C.

A Reactive Architecture for RoboCup Competition

1998 RoboCup-97: Robot Soccer World Cup I, pp. 434-442 inproceedings


Abstract: We illustrate PaSo-Team (The University of Padua Simulated Robot Soccer Team), a Multi-Agent System able to play soccer game for participating to the Simulator League of RoboCup competition. PaSo-Team looks like a partially reactive system built upon a number of specialized behaviors, just designed for a soccer play game and generating actions accordingly with environmental changes. A general description of the architecture and a guideline of main ideas is presented in the paper, whereas a more detailed description of actual implementation is given in the appendix.
Price, A., Jennings, A. & Kneen, J.

RoboCup97: An Omnidirectional Perspective

1998 RoboCup-97: Robot Soccer World Cup I, pp. 320-332 inproceedings


Abstract: RoboCup-97 proved to be a major learning curve. At RMIT University we took up the challenge of trying to build the most suitable robot platform for playing robot soccer. We investigated existing platforms, examined their strengths and weaknesses, and related each design to the needs of actual soccer players. We determined a set of criteria that we believe defines the needs of soccer playing robots of the future. Armed with this knowledge we set out to design a robot chassis that fulfilled, or at least, had the potential to fulfil as many of the desirable attributes as we could. Our approach is a long term one. It is very difficult to innovate in all areas of robotics at once. Unlike virtually all other teams we took a ground up approach, developing a unique, ideally suited robot platform first, giving a strong foundation to develop more sophisticated vision and control systems.
The development of omnidirectional sphere based technology at RMIT has produced a very lightweight practical omnidirectional drive for applications requiring rapid response, robust construction and isotropic maneuverability in an adversarial environment. The mechanism competed in the first world cup for Robot Soccer in Nagoya and won the inaugural Engineering Challenge award for innovation in design.
Price, A., Jennings, A., Kneen, J. & Kendall, E.

Learning, Deciding, Predicting: The Soccer Playing Mind

1998 RoboCup-97: Robot Soccer World Cup I, pp. 88-98 inproceedings


Abstract: In this paper we present an architecture which does not rely on any fixed hierarchy of agents in an attempt to remove dependencies and predictability from our system. We also present one method of simplifying the choice between available strategies by predicting the type of chassis used by the opposition, and therefore its limitation of movement. We consider the question of when are learning methods appropriate and in what areas are other, simpler approaches more efficient. We also develop coaching as a method of implanting simple maneuvers and tactical planning into our system. We consider the problem of how to distribute intelligence within the system, and the degree of independence and autonomy accorded to each player.
Riekki, J. & Röning, J.

Playing Soccer by Modifying and Combining Primitive Reactions

1998 RoboCup-97: Robot Soccer World Cup I, pp. 74-87 inproceedings


Abstract: We describe a novel approach for basing the actions of a mobile robot on the environment. The key idea is to produce continuously primitive reactions to all the important objects in the environment. Tasks are executed by modifying and combining these reactions. We have utilized this approach in playing soccer. We describe here the control system that controlled our players in the first RoboCup competition.
Rocher, S., Idasiak, V., Doncker, S., Drogoul, A. & Duhaut, D.

MICROB: The French Experiment in RoboCup

1998 RoboCup-97: Robot Soccer World Cup I, pp. 277-285 inproceedings


Abstract: In this paper, we present the ongoing research done in the field of robots playing "football". Several development aspects are discussed such as the hardware system, followed by the software implementation. We will highlight three softwares : the central decision system and the embedded software (used for the real robots small RoboCup league, and the simulation software (used for the simulation RoboCup league). So, we will show and discuss, for each kind of competition, the implemented behaviors.
Scerri, P.

A Multi-Layered Behavior Based System for Controlling RoboCup Agents

1998 RoboCup-97: Robot Soccer World Cup I, pp. 467-474 inproceedings


Abstract: We describe a multi-layered behavior based agent architecture which has been applied to the RoboCup domain. Upper layers are used to control the activation and priority of behaviors in layers below, and only the lowest layer interacts directly with the server. The layered approach seems to simplify behavior management. In particular it provides an approach for implementing high level strategic behavior in the soccer agents.
Shen, W.-M., Adibi, J., Adobbati, R., Cho, B., Erdem, A., Moradi, H., Salemi, B. & Tejada, S.

Autonomous Soccer Robots

1998 RoboCup-97: Robot Soccer World Cup I, pp. 295-304 inproceedings


Abstract: The Robocup 97 competition provides an excellent opportunity to demonstrate the techniques and methods of artificial intelligence, autonomous agents and computer vision. On a soccer field the core capabilities a player must have are to navigate the field, track the ball and other agents, recognize the difference between agents, collaborate with other agents, and hit the ball in the correct direction. USC's Dreamteam of robots can be described as a group of mobile autonomous agents collaborating in a rapidly changing environment. The key characteristic of this team is that each soccer robot is an autonomous agent, self-contained with all of its essential capabilities on-board. Our robots share the same general architecture and basic hardware, but they have integrated abilities to play different roles (goalkeeper, defender or forward) and utilize different strategies in their behavior. Our philosophy in building these robots is to use the least possible sophistication to make them as robust as possible. In the 1997 RoboCup competition, the Dreamteam played well and won the world championship in the middle-sized robot league.
Shinjoh, A.

RoboCup-3D: The Construction of Intelligent Navigation System

1998 RoboCup-97: Robot Soccer World Cup I, pp. 188-199 inproceedings


Abstract: RoboCup-3D is the attempt to build a virtual three-dimensional viewer for the RoboCup software league. Intelligent Navigation System(INS) is a basic concept of this RoboCup-3D, and it aims at creating a system which assists and navigates human cognition and activity in virtual three-dimensional space. Based on INS, we built SPACE, which is the first three-dimensional viewer for the RoboCup software league. In this paper, we are proposing the basic concept of this research and will show the possibility and the aims of future research of INS based on the analysis of SPACE.
Stone, P. & Veloso, M.

Using Decision Tree Confidence Factors for Multiagent Control

1998 RoboCup-97: Robot Soccer World Cup I, pp. 99-111 inproceedings


Abstract: Although Decision Trees are widely used for classification tasks, they are typically not used for agent control. This paper presents a novel technique for agent control in a complex multiagent domain based on the confidence factors provided by the C4.5 Decision Tree algorithm. Using Robotic Soccer as an example of such a domain, this paper incorporates a previously-trained Decision Tree into a full multiagent behavior that is capable of controlling agents throughout an entire game. Along with using Decision Trees for control, this behavior also makes use of the ability to reason about action-execution time to eliminate options that would not have adequate time to be executed successfully. This multiagent behavior represents a bridge between low-level and high-level learning in the Layered Learning paradigm. The newly created behavior is tested empirically in game situations.
Stone, P. & Veloso, M.

The CMUnited-97 Simulator Team

1998 RoboCup-97: Robot Soccer World Cup I, pp. 389-397 inproceedings


Abstract: The Soccer Server system provides a rich and challenging multiagent, real-time domain. Agents must accurately perceive and act despite a quickly changing, largely hidden, noisy world. They must also act at several levels, ranging from individual skills to full-team collaborative and adversarial behaviors. This article presents the CMUnited-97 approaches to the above challenges which helped the team to the semi-finals of the 29-team RoboCup-97 tournament.
Suzuki, S., Takahashi, Y., Uchibe, E., Nakamura, M., Mishima, C., Ishizuka, H., Kato, T. & Asada, M.

Vision-Based Robot Learning Towards RoboCup: Osaka University "Trackies

1998 RoboCup-97: Robot Soccer World Cup I, pp. 305-319 inproceedings


Abstract: The authors have applied reinforcement learning methods to real robot tasks in several aspects. We selected a skill of soccer as a task for a vision-based mobile robot. In this paper, we explain two of our method; (1)learning a shooting behavior, and (2)learning a shooting with avoiding an opponent. These behaviors were obtained by a robot in simulation and tested in a real environment in RoboCup-97. We discuss current limitations and future work along with the results of RoboCup-97.
Takaki, S.

The Reactive Motion Planning in the Passive Situation

1998 RoboCup-97: Robot Soccer World Cup I, pp. 428-433 inproceedings


Abstract: The HAARLEM1 is a team of the agents which is designed to play soccer. It is put emphasis on decision of behavior in the passive situation and it is set importance on the defense. The tactics of soccer are classified into three levels, individual tactics, group tactics and team tactics. The feature of the defense is embodied in the level of group tactics. The effectiveness of the tactics were demonstrated in the RoboCup-97, because the goal against was relatively low.
Tambe, M., Adibi, J., Al-Onaizan, Y., Erdem, A., Kaminka, G.A., Marsella, S.C., Muslea, I. & Tallis, M.

Using an Explicit Model of Teamwork in RoboCup

1998 RoboCup-97: Robot Soccer World Cup I, pp. 123-131 inproceedings


Abstract: Team ISIS (ISI Synthetic) successfully participated in the first international RoboCup soccer tournament (RoboCup'97) held in Nagoya, Japan, in August 1997. ISIS won the third-place prize in over 30 teams that participated in the simulation league of RoboCup'97 (the most popular among the three RoboCup'97 leagues. In terms of research accomplishments, ISIS illustrated the usefulness of an explicit model of teamwork both in terms of reduced development time and improved teamwork flexibility. ISIS also took some initial steps towards learning of individual player skills. This paper discusses the design of ISIS in detail, with particular emphasis on its novel approach to teamwork.
Veloso, M., Stone, P., Han, K. & Achim, S.

The CMUnited-97 Small Robot Team

1998 RoboCup-97: Robot Soccer World Cup I, pp. 242-256 inproceedings


Abstract: Robotic soccer is a challenging research domain which involves multiple agents that need to collaborate in an adversarial environment to achieve specific objectives. In this paper, we describe CMUnited the team of small robotic agents that we developed to enter the RoboCup-97 competition. We designed and built the robotic agents, devised the appropriate vision algorithm, and developed and implemented algorithms for strategic collaboration between the robots in an uncertain and dynamic environment. The robots can organize themselves in formations, hold specific roles, and pursue their goals. In game situations, they have demonstrated their collaborative behaviors on multiple occasions. The robots can also switch roles to maximize the overall performance of the team. We present an overview of the vision processing algorithm which successfully tracks multiple moving objects and predicts trajectories. The paper then focusses on the agent behaviors ranging from low-level individual behaviors to coordinated, strategic team behaviors. CMUnited won the RoboCup-97 small-robot competition at IJCAI-97 in Nagoya, Japan.
Verner, I.M.

The Value of Project-Based Education in Robotics

1998 RoboCup-97: Robot Soccer World Cup I, pp. 231-241 inproceedings


Abstract: This article deals with the educational values of the robotics tournaments as a form of project-oriented learning in engineering. The features of the project-oriented education, namely motivation, problem solving, time limits, interdisciplinary approach, team-work cooperation and applicable outcome, are discussed in relation to the Robotics area context. The RoboCup survey data about the participants, their motivation for participating in the program and preferences in the prospective summer school curriculum are summarized. Suggestions for incorporating educational activities in the RoboCup program are proposed.
Werger, B.B., Funes, P., Schneider-Fontán, M., Sargent, R., Witty, C. & Witty, T.

The Spirit of Bolivia: Complex Behavior Through Minimal Control

1998 RoboCup-97: Robot Soccer World Cup I, pp. 348-356 inproceedings


Abstract: The "Spirit of Bolivia" is a robotic soccer team which demonstrates minimally comprehensive team behavior. By this we mean that each member of the team makes progress towards team goals, and obstructs progress of the opponent, by interacting constructively with team-mates and in a sportsmanlike manner with opposing players. This complex behavior is achieved with simple on-board processors running very small behavior-based control programs; team behaviors are achieved without explicit communication. Externalization — the use of the environment as its own best model — and tolerance — a bias towards reducing the need for accurate information rather than attempting to recognize or correct noisy information — are the keys to robustness and sophistication of team behavior.
Yokota, K., Ozaki, K., Matsumoto, A., Kawabata, K., Kaetsu, H. & Asama, H.

Omni-Directional Autonomous Robots Cooperating for Team Play

1998 RoboCup-97: Robot Soccer World Cup I, pp. 333-347 inproceedings


Abstract: In order for multiple robots to accomplish a required task together, they need to communicate, organize themselves and cooperate. We have discussed these issues for the distribute autonomous robot system. The developed technologies are applied to the football game, in which the multiple mobile robots will maneuver and handle the ball towards the goal. The robots we use are holonomic, omni-directional mobile robots with vision and various sensors. They also have wireless LAN devices for communication. They can perform cooperation based on negotiation.
Zhang, Y. & Mackworth, A.K.

Using Reactive Deliberation for Real-Time Control of Soccer-Playing Robots

1998 RoboCup-97: Robot Soccer World Cup I, pp. 508-512 inproceedings


Abstract: Soccer meets the requirements of the Situated Agent approach and as a task domain is sufficiently rich to support research integrating many branches of Al. Reactive deliberation is a robot architecture that combines responsiveness to the environment with intelligent decision making. Under Reactive Deliberation, the robot controller is partitioned into a deliberator and an executor; the distinction is primarily based on the different time scales of interaction. A controller for our team entry in the Robocup97 Simulation League, UBC Dynamo97, has been developed using the Reactive Deliberation architecture.