HTTPS SSH

SARL Agent Controller for MAC Agents in the City 2017 - BASE

This project/package provides basic SARL controllers for the MAC Agents in City Contest. At this moment it only supports the 2017 edition.
It can be used as an initial set-up to build more advanced controllers.

The project relies on:

  1. The SARL Agents in City Middleware which provides SARL capacity and skill for teams to connect and play in the game simulator.
  2. The EISMASSim environment interface.
    • A Java library using the Environment Interface Standard (EIS) to communicate with the MASSim server that can be used with platforms which support the EIS.
    • Provides a more high-level access to the game sever than low-level JSON messages.
  3. The SARL-PROLOG-CAP project that provides a capacity and a skill for SARL agents to use SWI Prolog for implementing the knowledge base of the agents.

Some dummy controllers are provided as templates to build from.

IMPORTANT: For some support documentation/guides see:

  • A comprehensive set of instructions how to run SARL systems can be found here.
  • A FAQ list on using SARL+SWI via Maven and ECLIPSE IDE can be found here.

PREREQUISITES

  • Java Runtime Environment (JRE) and Java Compiler (javac) v1.8 (Sun version recommended)
  • Maven project management and comprehension tool (to meet dependencies, compile, package, run).
  • SARL (SRE Janus) execution engine:
  • The SARL Agents in City SARL Middleware.
    • Should be obtained automatically via Maven + Jitpack.
  • The EISMASSim a Java library using the Environment Interface Standard (EIS) to communicate with the MASSim server.
    • The JAR sources are provided under extras/ so they can be used to attach sources in ECLIPSE.
    • Comes with the game server. Using version 3.4 that comes with server massim-2017-1.7 (Sept 2017). Check here
    • Uses the eishub/EIS version 0.5 (sources also under extras/).
  • The SARL-PROLOG-CAP capacity+skill for SWI Prolog system:
    • Capacity (and skill) to allow SARL agents to have Prolog knowledge-bases.
    • Relies on JPL for the implementation to have SWI Prolog access in agents.
    • The right version (specified in the POM file) should be obtained automatically via JitPack:
      • From Bitbucket repo (less reliable): JitPack.
      • From Github clone (more reliable): JitPack
    • IMPORTANT: Please refer to the instructions and examples in the capacity+skill's page to set-up and use it in your SARL application.
  • The MASSIM Agents in City Game server: to run the game.
    • Server version 2017-0.7 that comes with massim package distribution massim-2017-1.7 (check release here).
    • You can also get the pre-pack binary from download section. You can also clone the version and run mvn clean package to build the massim package, including server, completely.

INSTALL, RUN and DEVELOP

You can run the SARL controller, either from ECLIPSE or from CLI (via Java or Maven), please refer to this instructions.

You can use the source JAR files for modules EIS and massim provided in extras/ to attah sources in ECLIPSE (their sources are not available via Maven).

So, to run the system you need to follow these general steps:

  1. Start MAC17 Game Server. For example, from server/ subdir:

    java -jar target/server-2017-0.7-jar-with-dependencies.jar --monitor 8001 -conf conf/Mexico-City-Test.json
    

    Please note the following important aspects of your game server configuration:

    • The configuration file (here, conf/Mexico-City-Test.json) makes a reference to the team configuration file at the bottom (e.g., conf/teams/A.json) which is the file containing all agents allowed to connect and with which id and password. These are the ones your system will use in your agent configuration file.
    • Configure the game server so that Prolog compatible terms are generated. In particular it is important to have names not starting with capital letters, as they will be undertood as variables.
      • Rename "2017-Mexico-City-Testto2017-mexico-city-test`.
      • Rename team names. Instead of team A, use team teamA.
  2. Start the SARL Controller, either via ECLIPSE or through the CLI (again, see general SARL instructions).

    • System will generally need a JSON configuration file for the game server.
    • By default, the JAR file built does not carry all dependencies as the compilation is too slow. Hence you need to execute via Maven execution plugin, which will run the default BootMAS class:

      mvn exec:java 
      mvn exec:java -Dexec.args=SuperSingleAgent -Dloglevel=4
      mvn exec:java -Dexec.args=SWISingleFullAgent -Dloglevel=4
      mvn exec:java -Dexec.args=BootMultiSWIAgents -Dloglevel=4
      

      If you run without arguments, it will list all available controllers and ask for one at console.

    • You can use grep inverse to get rid of all the printout of XML messages produced by the EI framework:

      mvn exec:java | grep -v -e \<.*> -e WARNING -e '^ sent' -e '^ received'

      This will filter out everything between < >, and any "sent" and "received" printout at start of line.

    • If you package the JAR with dependencies. then:

      java -jar target/sarl-agtcity-base-1.5.0.7.2-jar-with-dependencies.jar SWISingleFullAgent -Dloglevel=4
      

      or via the SARL booting class io.janusproject.Boot:

      java -cp target/sarl-agtcity-base-1.5.0.7.2-jar-with-dependencies.jar io.janusproject.Boot au.edu.rmit.agtgrp.agtcity.sarl.agents.dummy.SWISingleFullAgent -Dloglevel=4
      * You may want to grep to avoid the logging of all the XML messages: `grep -v \<.*\>`
      
  3. Start the MASSIM Simulation by just hitting ENTER in the Game Server console

  4. Enjoy! You should start seeing the agent reporting things in the console.
    • You can see the simulation on the web browser.

Developing SARL agents

You can check the example template agents (see below) that are in package au.edu.rmit.agtgrp.agtcity.sarl.agents.dummy to start your new development.
You will find there the process how SARL systems can connect to the game server and manipulate a team in the game.

If your solution will use Prolog as knowledgebase, we suggest carefully understanding how the SARL PROLOG CAP framework, which provides the skill to
manipulate a Prolog knowledgebase, works. Some initial code for a domain knowledge base is already provided in this base system, including substantial Prolog code to process all percepts.

We recommend using the JPL-based SWIJPL_KB_Prolog skill for the KB_Prolog capacity. It is simpler, direct to JPL and more expressive (e.g., can send Java objects to Prolog). Read the documentation in the SARL PROLOG CAP to understand how to use it to build queries, in particular the use of placeholders ? and the varios term types.

If you use the Mochalog-based SWI_KB_Prolog skill for the KB_Prolog capacity, then it is important to understand to read Mochalog to understand how to build queries using using @-placeholders @A, @I, and @S and the various query methods (prove, one solution, all solutions, iterators) provided.

EXAMPLE AGENTS

These are all "thin" agent controllers for the players in the game, but they should provide a solid base for developing more sophisticated agent systems.

All the agents can be run via the booting class BootMAS, which is the default execution class in the package JAR file. One has to give the name of the agent controller to start as an argument.

To change the log level, pass -Dloglevel=n option (FINE/DEBUG=4; INFO=3; WARNING=2; ERROR=1). Defaults to level 3.

The agent examples are:

  1. SuperSingleAgent: A simple central agent controlling all players in the simulator, receiving and processing the sensing from the environment, and then making all players navigate to random facilities. It uses the Java Percept Aggregator facility provided in the MW as percepts from players have a lot of shared content.

    • Class: au.edu.rmit.agtgrp.agtcity.sarl.agents.dummy.SuperSingleAgent
    • This is the default agent controller for BootMAS class so it can be run by just doing
    • To run:
      • Using Maven: mvn exec:java -Dexec.args=SuperSingleAgent -Dloglevel=4
      • Via main BootMAS class: java -jar target/sarl-agtcity-base-1.5.0.7.2-jar-with-dependencies.jar SuperSingleAgent
      • Via SARL booting class: java -cp target/sarl-agtcity-base-1.3.0.7.2-jar-with-dependencies.jar io.janusproject.Boot au.edu.rmit.agtgrp.agtcity.sarl.agents.dummy.SuperSingleAgent
      • Then, select the location where eismassimconfig.json server configuration file is located.
  2. SWISingleFullAgent: This is one single SARL agent controlling all players in the game and, importantly, using an SWI Prolog Knowledge Base via the JPL-based skill.

    • Class: au.edu.rmit.agtgrp.agtcity.sarl.agents.dummy.SWISingleFullAgent
    • To help understand how the SWI Knowledge Base is updated, every percept cycle the agent prints out some results from queries and then dumps its entire KB into file swiSingleFullAgent-<n>.pl, where <n> is the step number. The agent contains a few example uses of how to assert in the KB and query it. It also contains simple logic to continously select a random destination and go there. For achieving this, both SARL (emitting and handling event E_MoveRandomly) and SWI Prolog (for choosing the destination) is used.
    • To run this agent: mvn exec:java -Dexec.args=SWISingleFullAgent -Dloglevel=4, and then select the location where eismassimconfig.json server configuration file is located.
  3. Multi Agent System: This is a multi SARL agent team started with agent BootMultiSWIAgents which starts a set of agents of type SWISingleFullAgent. Each of them is assigned a configuration file to control a set of player in the simulation. To do so, the BootMultiSWIAgents overrides the default values of SWISingleFullAgent to provide authentication files directly as arguments to the initialization.

    • Class: au.edu.rmit.agtgrp.agtcity.sarl.agents.dummy.BootMultiSWIAgents
    • This is a perfect example of versatility where a SARL agent can be given the task to control all players or just a subset.
    • To run this agent:
      • Using Maven: mvn exec:java -Dexec.args=BootMultiSWIAgents -Dloglevel=4
      • Using the main class BootMAS (after compiling with all dependencies): java -jar target/sarl-agtcity-base-1.3.0.7.2-jar-with-dependencies.jar BootMultiSWIAgents
    • Then, select the location where eismassimconfig.json server configuration file is located.

For general links check here.

PROJECT CONTRIBUTORS

  • Sebastian Sardina (Project leader and contact - ssardina@gmail.com)
  • Matthew McNally (first version of SWI-based agent via Mochalog)

LICENSE

This project is using the GPLv3 for open source licensing for information and the license visit GNU website (https://www.gnu.org/licenses/gpl-3.0.en.html).

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.