Common Hybrid Agent Platform
SourceForge.net Logo

Many-Agent Systems

The Common Hybrid Agent Platform (CHAP) is an open source project sponsored by Almende BV, a research company in Rotterdam, The Netherlands. CHAP serves as a framework for developing multi-agent systems (MAS) that bring together artificial entities and humans. Self-organization is something that can be observed in human behavior, and the CHAP agent systems attempt to improve the self-organization amongst people by exploring the accelerated interactions of communicating agents.

Goals of CHAP

Besides some of the standard functionalities of frameworks like JADE and Cougaar, CHAP aims to provide:

  • Multiple programming language support (agents can be programmed in Java, C, ObjectiveC, etc)
  • Various technologies to facilitate and enhance human-agent interaction
  • Agent Visualization tools

Subprojects

Abbey

Abbey is an experiment in low-level code (written in C) which attempts to tackle the issues of concurrency and muliprocessing by exploring alternative approaches to coordination. Workers called "monks" are enlisted to carry out atomic computational tasks and coordinate the data flow amongst the tasks through the use of channels.

CAL (Communicating Agent Library)

CAL, written in Objective C using GNUStep), is the coordination logic that underlies the DEAL project (Distributed Engine for Advanced Logistics). Agents representing trucks and packages interact asynchronously and make deals by performing auctions.

Groovy Actors

Groovy Actors unites the Groovy dynamic language with java.util.concurrent so that a large population of autonomous Groovy scripts can be run simultaneously within one JVM. The actors communicate amongst themselves using Closures. Key features: Asterisk and Jabber integration for human-agent interaction.

Emerge

Emerge is a Java-based platform for rapidly developing systems consisting of large numbers of simple autonomous agents which communicate asynchronously with each other. Key features: Agent visualizations, Face Agents.