Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

GAMMS: Graph based Adversarial Multiagent Modeling Simulator

Rohan Patil, Jai Malegaonkar, Xiao Jiang, Andre Dion, Gaurav S. Sukhatme, Henrik I. Christensen

TL;DR

GAMMS addresses the need for scalable, accessible multi-agent simulation on graph-structured domains by introducing a graph-based, integration-friendly framework that supports diverse agent policies, including learning-based and human-in-the-loop approaches. The system emphasizes five core objectives—scalability, ease of use, integration-first architecture, fast visualization feedback, and real-world grounding—implemented via modular components (Graph, Sensor, Agent, Recorder, Logger, Visualization) and a single, authoritative context. It offers a Python-native, minimal-interface design with an OSMnx-driven pipeline to convert real-world maps into graphs, enabling large-scale experiments on standard hardware. Through Grid World and Capture-the-Flag demonstrations on real map data, including aerial and ground agents, GAMMS showcases rapid prototyping, visualization, and extensible integration, positioning it as a practical tool for advancing multi-agent planning, adversarial modeling, and embodied AI research.

Abstract

As intelligent systems and multi-agent coordination become increasingly central to real-world applications, there is a growing need for simulation tools that are both scalable and accessible. Existing high-fidelity simulators, while powerful, are often computationally expensive and ill-suited for rapid prototyping or large-scale agent deployments. We present GAMMS (Graph based Adversarial Multiagent Modeling Simulator), a lightweight yet extensible simulation framework designed to support fast development and evaluation of agent behavior in environments that can be represented as graphs. GAMMS emphasizes five core objectives: scalability, ease of use, integration-first architecture, fast visualization feedback, and real-world grounding. It enables efficient simulation of complex domains such as urban road networks and communication systems, supports integration with external tools (e.g., machine learning libraries, planning solvers), and provides built-in visualization with minimal configuration. GAMMS is agnostic to policy type, supporting heuristic, optimization-based, and learning-based agents, including those using large language models. By lowering the barrier to entry for researchers and enabling high-performance simulations on standard hardware, GAMMS facilitates experimentation and innovation in multi-agent systems, autonomous planning, and adversarial modeling. The framework is open-source and available at https://github.com/GAMMSim/GAMMS/

GAMMS: Graph based Adversarial Multiagent Modeling Simulator

TL;DR

GAMMS addresses the need for scalable, accessible multi-agent simulation on graph-structured domains by introducing a graph-based, integration-friendly framework that supports diverse agent policies, including learning-based and human-in-the-loop approaches. The system emphasizes five core objectives—scalability, ease of use, integration-first architecture, fast visualization feedback, and real-world grounding—implemented via modular components (Graph, Sensor, Agent, Recorder, Logger, Visualization) and a single, authoritative context. It offers a Python-native, minimal-interface design with an OSMnx-driven pipeline to convert real-world maps into graphs, enabling large-scale experiments on standard hardware. Through Grid World and Capture-the-Flag demonstrations on real map data, including aerial and ground agents, GAMMS showcases rapid prototyping, visualization, and extensible integration, positioning it as a practical tool for advancing multi-agent planning, adversarial modeling, and embodied AI research.

Abstract

As intelligent systems and multi-agent coordination become increasingly central to real-world applications, there is a growing need for simulation tools that are both scalable and accessible. Existing high-fidelity simulators, while powerful, are often computationally expensive and ill-suited for rapid prototyping or large-scale agent deployments. We present GAMMS (Graph based Adversarial Multiagent Modeling Simulator), a lightweight yet extensible simulation framework designed to support fast development and evaluation of agent behavior in environments that can be represented as graphs. GAMMS emphasizes five core objectives: scalability, ease of use, integration-first architecture, fast visualization feedback, and real-world grounding. It enables efficient simulation of complex domains such as urban road networks and communication systems, supports integration with external tools (e.g., machine learning libraries, planning solvers), and provides built-in visualization with minimal configuration. GAMMS is agnostic to policy type, supporting heuristic, optimization-based, and learning-based agents, including those using large language models. By lowering the barrier to entry for researchers and enabling high-performance simulations on standard hardware, GAMMS facilitates experimentation and innovation in multi-agent systems, autonomous planning, and adversarial modeling. The framework is open-source and available at https://github.com/GAMMSim/GAMMS/
Paper Structure (9 sections, 6 figures)

This paper contains 9 sections, 6 figures.

Table of Contents

  1. Introduction
  2. GAMMS Objectives
  3. Related Works
  4. Simulator Architecture
  5. Design Principles and Choices
  6. Simulator Components
  7. Overall Flow and Construction of the Game Loop
  8. Examples
  9. Discussion & Future Work

Figures (6)

  • Figure 1: Central Park, NYC OSM on the left. GAMMS graph for Central Park on the right.
  • Figure 2: The Figure shows the current overall architecture of GAMMS.
  • Figure 3: Overall flow and construction of a Game using GAMMS
  • Figure 4: Part of the grid world created by GAMMS.
  • Figure 5: Part of La Jolla OSM converted into a GAMMS graph for the Capture the Flag scenario. Red and blue areas show the territories of the respective teams. The big sqares show the flag positions.
  • ...and 1 more figures