Holistic Digital Human Models in Gazebo: A Case Study on Agricultural Workflows

Abstract

The agricultural sector is undergoing digital transformation due to modern automation, robotics, sensory, and simulation technologies. This research explores a digital human models (DHMs) usage in the Gazebo virtual environment to enhance agricultural workflows, human-robot interaction, and safety. We propose a framework to simulate typical agricultural scenarios, such as field mapping, harvesting efficiency control, crop inspection, obstacle avoidance, and theft detection. Farm workers’ DHMs model interactions with mobile autonomous systems, stationary sensors and sensory networks. The DHMs are supplied with generic and task-specific animations; the latter include such activities as crop harvesting or field inspection. The simulation environment features agricultural settings with dynamic obstacles and predefined work zones. Assessment metrics, such as task completion time or an obstacle avoidance rate, allow measuring performance of each scenario. Results of preliminary simulation of the proposed simplified scenarios in the Gazebo simulator demonstrated a high potential of DHMs and Gazebo to optimize agricultural workflows and improve human-robot interaction. This research provides a foundation for using simulation technologies to address real-world agricultural issues.