Heuristic Approaches for Coordinating Collaborative Heterogeneous Robotic Systems in Harvesting Automation with Size...

Area of Science:

• Robotics
• Artificial Intelligence
• Operations Research

Background:

• Coordinating heterogeneous robotic systems in agriculture, especially for harvesting, is complex due to task allocation, routing, and scheduling needs.
• Real-time sensing and rapid coordination updates are crucial for efficient agricultural operations.

Purpose of the Study:

• To develop and compare three heuristic approaches for multi-agent coordination in agricultural harvesting automation.
• To address challenges of robot size constraints and optimize task completion time.

Main Methods:

• Primal-dual workload balancing inspired by combinatorial optimization.
• Greedy task assignment with iterative local optimization.
• Large Language Model (LLM)-based constraint processing via prompt engineering.

Main Results:

• All three heuristic approaches yielded feasible solutions for agricultural harvesting automation.
• The methods demonstrated reasonable solution quality in extensive simulations.
• The primal-dual and greedy methods focused on workload balancing and route optimization, while the LLM approach handled constraints.

Conclusions:

• The developed heuristic methods show significant potential for real-world agricultural applications.
• These approaches can be adapted to variations of multi-robot coordination problems.
• The study provides valuable insights into solving complex coordination challenges with heterogeneous multi-robot systems.