aamas_mcp_2025

# AAMAS Maritime Shipping Competition 2025 - Team 7 This repository contains our team's solution for the **Maritime Shipping Competition (MSC)** at AAMAS 2025, part of the 24th International Conference on Autonomous Agents and Multiagent Systems. ## 🚢 Competition Overview The Maritime Shipping Competition challenges participants to develop intelligent software agents in Python that manage a simulated maritime shipping company. The competition focuses on two main aspects: - **Trade Auction Bidding**: Developing strategic bidding algorithms for cargo transportation contracts - **Fleet Scheduling**: Optimizing vessel routes and schedules for maximum efficiency and profit Participants compete to create the most innovative and effective approaches to these maritime logistics challenges, with prizes awarded for both performance and creativity. ### 🏆 Competition Website For more details about the competition, visit: [Maritime Shipping Competition AAMAS 2025](https://jbuerman.github.io/Maritime-Shipping-Competition-AAMAS2025/tournament/) ## 🎯 Our Approach Our team has developed a sophisticated multi-agent system that combines: ### 1. **Advanced Optimization Techniques** - **Constraint Programming**: Using OR-Tools CP-SAT solver for optimal vessel-trade assignments - **Greedy Algorithms**: Fast heuristic approaches for real-time decision making - **K-Best Solutions**: Exploring multiple high-quality solutions for robust performance ### 2. **Intelligent Bidding Strategies** - Dynamic bid calculation based on route profitability - Risk assessment considering vessel availability and scheduling conflicts - Competitive analysis and market adaptation ### 3. **Efficient Route Planning** - Pickup and Delivery Vehicle Routing Problem (PDVRP) formulation - Time window optimization for trade scheduling - Cost allocation mechanisms for fair profit distribution ## 🔧 Key Features - **Multi-Strategy Implementation**: Multiple agent variants (greedy, k-best, bidding-optimized) - **Precomputed Route Cache**: Optimized distance calculations for faster decision making - **Comprehensive Testing Suite**: Extensive validation and performance analysis - **Modular Design**: Clean separation of concerns for easy maintenance and extension ## 📁 Repository Structure ``` ├── Agents.py # Core solver implementation with CP-SAT optimization ├── group7.py # Main team submission with advanced scheduling algorithms ├── greedy.py # Greedy heuristic implementation ├── kbest.py # K-best solutions algorithm ├── kbest_bid.py # Bidding strategy with k-best optimization ├── utils.py # Utility functions and helper methods ├── run_experiments.py # Performance testing and benchmarking ├── test.py # Unit tests and validation ├── example.py # Basic example implementation ├── precomputed_routes.pickle # Cached route calculations ├── mable_resources/ # Competition framework resources ├── *.ipynb # Jupyter notebooks for analysis and development └── requirements.txt # Project dependencies ``` ## 🚀 Getting Started ### Prerequisites ```bash pip install -r requirements.txt ``` ### Running the Agent ```bash # Basic test run python test.py # Run experiments python run_experiments.py # Execute with shell script ./run.sh ``` ### Key Dependencies - **OR-Tools**: Constraint programming solver - **MABLE Framework**: Competition simulation environment - **Loguru**: Advanced logging - **Marshmallow**: Data serialization - **NumPy/Pandas**: Data processing (as needed) ## 🧠 Technical Highlights ### 1. **Constraint Programming Model** Our core solver uses CP-SAT to formulate the vessel scheduling problem as a mixed-integer programming problem, considering: - Trade assignment constraints - Time window requirements - Vessel capacity limitations - Travel time calculations ### 2. **Cost Allocation Algorithm** Sophisticated cost sharing mechanism that fairly distributes: - Travel costs among trades sharing vessel routes - Operational costs for port activities - Idle time penalties ### 3. **Bidding Intelligence** Strategic bidding that considers: - Route profitability analysis - Competitive market dynamics - Risk assessment based on schedule feasibility ## 🎯 Innovation Elements - **Shared Cost Allocation**: Novel approach to distribute travel costs among multiple trades - **Hybrid Optimization**: Combining exact methods (CP-SAT) with fast heuristics - **Adaptive Bidding**: Dynamic strategy adjustment based on market conditions - **Precomputed Optimization**: Smart caching for real-time performance ## 📊 Performance Our solution demonstrates: - Efficient handling of large-scale scheduling problems - Competitive bidding performance - Robust optimization under various market conditions - Scalable architecture for different fleet sizes ## 🤝 Team Information **Team 7** - Participating in AAMAS 2025 Maritime Shipping Competition ## 📄 License This project is developed for the AAMAS 2025 Maritime Shipping Competition. Please refer to the competition guidelines for usage terms. --- *For questions about our approach or collaboration opportunities, please feel free to reach out!*