The Present and Future of Soccer-Robot: The Present and Future of Soccer-Robot Institute of Artificial Intelligence and Robotics Northeastern University, China Xu Xinhe


Content: Content Introduction to robot soccer Reasoning model of strategy system of robot soccer The goal and future of robot soccer


Slide3: Introduction to robot soccer


Why it’s robot soccer?: Why it’s robot soccer? Many kinds of sport can be played by robot, but robot soccer is more and more popular and developed very rapidly. Why? Soccer is the first sport in the world and very attractive to young people. Some foreseeing scientists and pioneers have done valuable work.


We have to mention: We have to mention 1992，Prof. A1an Mackworth (University of British Columbia, Canada) 1993，Prof. Minoru Asada（浅田埝）、Hiroaki Kitano（北野宏明）and Yasuo Kuniyoshi set up RoboCup game


Slide6: RoboCup – The Robot World Cup Initiative Japan， 1997： Category of soccer： Simulation (11:11) Small Size (5:5) Middle Size (4:4) SONY Legged Humanoid League Junior Rescue Robot Soccer World Cup


Slide7: FIRA – Federation of International Robot Soccer Association KAIST,Korea,1996： Category： MiroSot(3:3, 5:5) NaroSot(5:5) RoboSot KheperaSot SimuSot(5:5, 11:11) Benchmark Teat Robot Soccer World Cup


The Common View: The Common View Robot soccer is a grand challenge of robotics and Artificial Intelligence A desired research and competition platform of Artificial Intelligence


Artificial Intelligence: Artificial Intelligence Many kinds of definition In short, artificial intelligence is the understanding and implementation of human intelligence, including apperceive thinking and decision-making action and expression


Slide10: initiative apperceive thinking decision-making action Understanding human soccer (single player)


Slide11: communication Understanding human soccer (multi players) I II III IV initiative apperceive thinking decision-making action


Slide12: Robot Soccer Implement


Slide13: RoboCup Simulation League


Slide14: communication I II III IV initiative apperceive thinking decision-making action Robot Soccer Implement


Slide15: Vision-Based Remote Brainless Multi-Robot System


Vision-Based Remote Brainless Multi-Robot System: Vision-Based Remote Brainless Multi-Robot System


Mirosot Middle League of Fira: Mirosot Middle League of Fira


Slide18: communication I II III IV initiative apperceive thinking decision-making action Robot Soccer Implement


Slide19: Vision Based Information Express Vision-Based Brain-on-Board Multi-Robot System


Vision-Based Brain-on-Board Multi-Robot System: Vision-Based Brain-on-Board Multi-Robot System


Slide21: Small Size Robot League of Robocup


Slide22: communication I II III IV initiative apperceive thinking decision-making action Robot Soccer Implement


Slide23: Respective Apperceive Primary Decision Communication Ultimate Decision Action Robot-Based Multi-Robot System


RoboSot League of Fira: RoboSot League of Fira


Middle Size Robot League of Robocup: Middle Size Robot League of Robocup


Slide26: Wheeled robot Car-like tracklayer Legged robot multi-leg (Sony AIBO) biped (Humanoid) Advanced Implement


Sony AIBO Robot League: Sony AIBO Robot League


Humanoid League: Humanoid League


Humanoid League (Robocup): Humanoid League (Robocup)


HuroSot League of Fira: HuroSot League of Fira


Slide31: Reasoning Model of Strategy System of Robot Soccer


System Structure Analysis: System Structure Analysis


Slide33: Playground Digital Image Decision-making Subsystem Car Controller Pattern Recognition R3 R2 R3’ R1’ R2’ R1 m1 m2 com com com com Cars Subsystem Ball The structural block diagram of robot-soccer system Communication Subsystem Vision Subsystem


The imaginative explanationof decision making: The imaginative explanation of decision making


It is obvious that: It is obvious that The blind coach can receive no image information but digital position information. As a result, he can not carry out imaginative thinking, but with a series of logical reasoning on the basis of data and symbols processing. Therefore, when we formalize the expert knowledge and reasoning process, we must carefully abstract each link of the inference chain with the aid of the thinking expansion of the blind coach.


Six-step Reasoning Model: Six-step Reasoning Model Step 1: Pretreatment of input information (distance, velocity and relative angle calculation etc.) . Step 2: Judgement of offensive and defensive posture. Step 3: Determination of battle formation and role assignment. Step 4: Determination of the target position. Step 5: Path planning. Step 6: Calculation of the wheels’ moving direction, velocity and displacement.


Generally speaking: Generally speaking step 1,5,6 --- deductive reasoning can be performed through mathematical model step 2,3,4 --- production reasoning can only be fulfilled with expert system


Slide38: The content of each step Step 1: Pretreatment of input information Includes the calculation of each entity linear velocity angular velocity relative distances and angles in order to judge which side or player is controlling or most possible to control the ball


Slide39: Example: Situation in moment


Slide40: Input Matrix: X Y θ R1 72 41 193 R2 71 75 167 R3 4 65 360 ER1 106 49 272 ER2 110 84 155 ER3 145 73 90 B 88 71 ? Missing a lot of information! Recover it by Pretreatment


Slide41: Step 2: Judgement of offensive and defensive posture This step is the important premise of tactics-making in the field It is a very complex problem since situation in the field change all the time it is impossible to know the opposite intention the posture is a fuzzy concept no common point view The content of each step


Slide42: Step 3: Determination of battle formation and role assignment Some team battle formations (attack to right) The content of each step


Slide43: Offensive situation ↓ Offensive battle formation 1-1-1 or 1-0-2 ↓ Role assignment : Main attacker (in good condition) Assistant attacker Goalie


Slide44: Defensive situation ↓ Defensive battle formation 1-2-0 ↓ Role assignment Main guarder (block the ball) Assistant guarder goalkeeper


Slide45: Step 4: Determination of the target position. The main attacker should arrive at some point of the prolong line between opposite goal and ball, and go forward with the ball. The defender should stand in the line which connects the ball and own goal. The other players should arrive at the pointed position quickly according to the strategy. The content of each step


Slide46: Step 5: Path planning There are many advancing routes from the current position to the target position The content of each step


Slide47: One method of path planning


Slide49: Step 6: Calculation of the wheels’ velocity and displacement. The mapping relation from path to wheels’ velocity and displacement is the basic algorithm in kinematics. The existent problem is not only to give the set point of the left and right wheels, but also to notice that the car is always in the transition process of starting and stopping. If we use stepper motor in stead of direct-current motor to drive the car, the calculation for pulse quantity would be easier. The content of each step


Slide50: Combine Step 4、5、6 into Action design, such as Basic Action To_position Turn_angle Skilled Action Shoot_ball Block_ball Block_man Goal_defense Conner_processing …… In practice


Slide51: Input information from Computer Vision Subsystem Pretreatment of input information Role assignment Action assignment Action complementation Command on wheels’ velocity Send to communication subsystem Structure of Decision-making Subsystem


Slide52: The Six-step Reasoning Model provides a basic frame for the development of the Strategy system of robot-soccer.


Slide53: The Goal and Future of Robot Soccer


The Goal --- The Dream: The Goal --- The Dream By mid-21st century, a team of fully autonomous humanoid robot soccer players shall win the soccer game, comply with the official rule of the FIFA, against the winner of the most recent World Cup.


Landmark Project: Landmark Project A significant long term goal A landmark project as well as a standard problem The formidable challenge will be shared by robotics and AI community for next 50 years May sound overly ambitious given the state of the art technology today


Is it feasible ?: Is it feasible ? It took only 50 years From Write Brother’s first aircraft to Apollo mission to send men to the moon and safely return From the invention of digital computer to the Deep Blue, which beat human champion Garry Kasparov in chess It is hard to image the progress of science and technology in the next 50 years


Meaning of the dream : Meaning of the dream A series of technical breakthrough is necessary Wide range of technologies can be integrated and examined Researchers can get together and evaluate research progress Being used for integrated project-oriented education


Wide Range of Technologies : Wide Range of Technologies Design principles of autonomous agents Multi-agent collaboration Strategy acquisition Real-time reasoning Robotics Real-time sensor-fusion Reactive behavior Context recognition Strategic decision-making Motor control Intelligent robot control And many more


Good beginning : Good beginning


Good beginning: Good beginning


Long way to go : Long way to go


Long way to go: Long way to go


Slide63: Thank you for attention The End