Download presentation
Presentation is loading. Please wait.
Published byJonathan Horton Modified over 9 years ago
1
Vermelding onderdeel organisatie September 12, 2015 1 Intro Multi-Agent Systemen Multi-Agent Systemen Koen Hindriks, Birna van Riemsdijk Man-machine interaction - Mediamatics
2
September 12, 20152 Teachers Faculty dr. Koen Hindriks dr. Birna van Riemsdijk Student Assistants Jurian van Dalfsen Marijn Goedegebure Michiel Hegemans Vincent Koeman Camiel Steenstra Marieke van der Tuin
3
Course setup, materials & instructions Thursday: Lecture 8:45-10:30 (EWI-Collegezaal A) Friday: Tutorial 13:45-15:30 (Drebbelweg) Work on and assistance with assignments Blackboard Course information 4x homework assignments Feedback Announcements Book: Learn Prolog Now! Additional materials Discussion forum September 12, 20153
4
Grading 4 Practical assignments In pairs! First deadline: Feb 25th Graded with sufficient/insufficient 0.25 points bonus per sufficient assignment No retake, no deadline extensions Written exam Final grade Exam grade + bonus points (not for resits) For details: see blackboard! 4
5
Schedule Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Introduction to Agents & Basics of Prolog Prolog: Lists, Arithmetic Prolog: Cuts & Negation GOAL: Mental states of agents GOAL: Actions & macros GOAL: Environments & perception, and modules GOAL: Multi-agent systems & communication September 12, 20155
6
Multi-Agent Systemen – Overview First part Brief intro to Agents Second part Prolog as a Knowledge Representation Technology 9/12/20156
7
Multi-Agent Systems – Perspective 9/12/20157 Project Multi-Agent Systems: CTF Competition in UT2004 Control a team of bots by means of a multi-agent system. Compete at the end of the project. Course Multi-Agent Systems: Learn to program a multi-agent system Develop logic-based agents programs: Apply reasoning technology (Prolog) Write agent programs (GOAL) Hands-on experience by various programming assignments.
8
WHAT KIND OF SOFTWARE DO WE NEED? Bots that proactively need to choose actions in real-time and in response to events that happen in the gaming environment 9/12/20158
9
Programming for Challenging Environments Responsive software… Programs that are flexible, change behaviour when needed Programs that are reactive, respond to events and changes in a timely manner A robust behavioural layer Proactive software… Programs that are goal- directed, pursue a desired state of affairs Programs that act strategically, are able to coordinate and to compete with other agents …aimed at achieving goals 9/12/20159 We need…
10
September 12, 201510 Software Agents Flexible Robust Reduced coupling Complexity Decentralization http://www.ida.liu.se
11
September 12, 201511 Intelligent Agents Reactive – respond in timely manner to change Proactive – (persistently) pursues multiple, explicit goals over time Social – agents need to interact and perform e.g. as a team Agents are high-level (or rational) action selection mechanisms Autonomous means: Agent independently makes its own decisions Autonomous Agents
12
September 12, 201512 Situated agents: actions & percepts Choose an action Percepts Action environmentagent Populated world
13
September 12, 201513 Situated agents: actions & percepts ? Percepts Action environmentagent Populated world
14
WHAT IS A CHALLENGING ENVIRONMENT? 9/12/201514
15
Autonomous Unmanned Vehicles 9/12/201515 Application: Next Generation Vehicles Competitions: See http://www.auvsifoundation.org/http://www.auvsifoundation.org/ Intelligent Ground Vehicle Competition International Aerial Robotics Competition Autonomous Surface Vehicle Competition Autonomous Underwater Vehicle Competition Goal: a world where human lives are protected by and enhanced by the regular use of robotic technologies. Princeton Autonomous Vehicle Google Driverless Car TU-Berlin helicopter-based aerial robot
16
Application: Future Rescue Systems, Training Crisis Management 9/12/201516 Competitions: See http://www.robocuprescue.org/http://www.robocuprescue.org/ Urban Search and Rescue Robot Competitions Rescue Simulation League Rescue Robot League Goal: to increase awareness of the challenges involved in search and rescue applications, and to provide objective evaluation of robotic implementations iRobot PackBot in WTC
17
Application: Household robots, … Soccer / Sports 9/12/201517 Competitions: See http://www.robocup.org/http://www.robocup.org/ RoboCup Soccer Leagues: Simulation League, Small Size Robot League, Middle Size Robot League, Humanoid League, Standard Platform League Goal: By the year 2050, develop a team of fully autonomous humanoid robots that can win against the human world soccer champion team. Aldebaran Nao Small Medium Simulation Standard Humanoid TU Delft MMI
18
September 12, 201518 This Course Learn programming language (GOAL) to program action selection mechanism of intelligent agents GOAL uses the logic programming language Prolog as its knowledge representation language
19
September 12, 201519 Proactive & reactive agents: goals & events agent Percepts Action environment Populated world events actionsgoals
20
September 12, 201520 Coping with the environment: plans & beliefs agent Percepts Action environment Populated world events actionsgoals plans beliefs
21
Organisation-based design Agent-oriented programming (AOP) Object-oriented languages (OOP) Knowledge-oriented languages (KR) Procedural languages Assembler Machine language September 12, 201521 Evolution of Programming abstraction time e.g. GOAL e.g. Prolog Bots that proactively need to choose actions in real-time and in response to events that happen in the gaming environment
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.