Multi-Agent Auction-Based Task Allocation A research-grade simulation of autonomous multi-agent coordination using decentralized auction mechanisms.

Overview

This project implements a clean, modular, and academically relevant simulation of multi-agent systems (MAS) where a team of autonomous mobile agents cooperatively allocates and completes tasks in a 2D environment.

Each agent makes local, independent decisions, communicates through lightweight messages, and participates in auction-based task allocation (Contract Net style), one of the core paradigms in distributed AI and robotics.

The system demonstrates:

Decentralized coordination

Local decision-making

Emergent load balancing

Distributed task allocation

Robust behavior without a central controller

This repository is designed as a portfolio-quality research project suitable for PhD applications in Autonomous Systems, Robotics, AI, and Multi-Agent Coordination.

Key Features

Autonomous agents with local perception

Auction-based task allocation (CFP, BID, AWARD)

Dynamic task progress and completion

Message-based communication

Modular architecture (Agent, Task, World, Message)

Fully reproducible simulation example

Real-World Relevance

This architecture reflects the same principles used in:

Multi-robot exploration

Drone swarm mapping

Autonomous delivery fleets

Smart factory coordination

Decentralized energy management

Traffic-aware autonomous vehicles

The implementation closely matches concepts used in research on:

Contract Net Protocol

Market-based Multi-Robot Task Allocation (MRTA)

Distributed Planning

Swarm Robotics

Agent-Based Modeling

Repository Structure

multi-agent-task-allocation/ src/ world.py Environment, auctions, simulation logic agent.py Autonomous agents task.py Task entities message.py Inter-agent message model simulation.py Demo environment and runner examples/ run_demo.py Runnable simulation example notebooks/ analysis.ipynb Optional Jupyter analysis README.md requirements.txt LICENSE .gitignore

Running the Demo

Clone the repository: git clone https://github.com/alizangeneh/multi-agent-task-allocation

Navigate into the directory: cd multi-agent-task-allocation

Run the demo: python examples/run_demo.py

About the Author

Ali Zangeneh Software Engineer • AI/ML Researcher • Robotics & Multi-Agent Systems Enthusiast

Email: [email protected]

GitHub: https://github.com/alizangeneh

ORCID: https://orcid.org/0009-0002-5184-0571

Research interests:

Multi-Agent Reinforcement Learning

Autonomous Robotics and Swarm Intelligence

Explainable and Trustworthy AI

Optimization and Intelligent Decision-Making

Foundation Models for Robotics and Agents

This project is part of my academic portfolio for PhD applications in Europe and the UK.

License

MIT License — Free to use, modify, and extend.