Welcome to CHAP. CHAP is a software development toolkit developed by Almende. It offers generic multi-agent solutions for a myriad of problems.
CHAP includes software development tools, models and concrete solutions resulting from all Almende research projects. All these components are generic, so they may be tailored and diversified to suit different applications.
All Almende research is centered around the concept of self-organization. Self-organization is the process by which complexity is handled in a system without central guidance from an external source. Self-organization is a topic addressed by many academic domains: physics, biology, mathematics, information theory, economics and sociology.
Almende believes that the best way of achieving self-organization in human society is with the aid of adaptive, robust and scalable software; more specifically, through multi-agent technology. CHAP can be seen as a library or a toolkit full of agent-based solutions. Years of experience in research and development at Almende are collected within CHAP.
A fully functioning self-organized system makes use of the autonomy of individuals to achieve its goals, thereby creating a network of people who can act independently. Software agents can be used to represent and aid each person in this network. Agents may also represent an object, an organization, a concept or a specific goal; any action or object in the real world.
Agents function in hypertime. They can communicate and coordinate much faster than humans can. An agent functions like a personal assistant to the person or object it is assigned to, working to achieve the goals for their reallife counterpart. Individual agents can facilitate quick and frequent communication, help making and keeping appointments, negotiate deals and track activities.
There is no central structure which is implemented from above. Each agent determines its own actions and negotiates and coordinates with other agents in the system. They can adapt to new situations and information, which is important if the software is to be implemented in a real, dynamic environment. Agents are designed to ask for feedback if they are confronted with a problem they cannot solve by themselves. They are not programmed for each eventuality, but to learn from experience and (implicit) feedback.