MAX

MAX (for ‘Multi-Agent eXperimenter’) is a framework for agent-based simulation and full agent-based applications building. It was authored by Önder Gürcan in 2018 and has been in continuous development ever since at the CEA. The objective of MAX is to make rapid prototyping of industrial cases and to carry out their feasibility analysis in a realistic manner.

MAX is particularly well suited for modeling open, distributed and intelligent systems developing over time. It embeds the Agent/Environment/Role (AER) organizational model.

MAX is a deployed platform that enables developers to build models in the safety of their own environment but provides them with the capability to more realistically test their algorithms and improve execution quality. By recreating the dynamics of the system at an agent level, the impact of actions executed by an algorithm under different scenarios can be tested and analyzed in granular detail.

MAX has extensive documentation and tutorials. It also comes with a model library, a collection of pre-written simulations that can be used and modified. These simulations, for the moment, address content areas in the network research including distributed ledger technologies research.

MAX is the next generation of the multi-agent frameworks including MadKit and Repast. MAX runs on the Java virtual machine, so it works on all major platforms (MacOS, Windows, Linux, et al). It can run as a desktop application and a command line application.