Autonomous collaborative mobile robot for greenhouses: Design, development, and validation tests

Abstract

This paper describes the development of a mobile agricultural robot capable of performing high-capacity transport tasks within greenhouses in presence of people or other agricultural machines. The main objective is to provide the robot with enough technology to work collaboratively with nearby human workers. In addition, the robot must also be able to transport 100 kilograms in a safe way over uneven terrain, a characteristic not usually found in existing greenhouse robots. This is important to ensure the sustainability of intensive greenhouse cultivation, as it is essential to allow more flexible use of robots when adapting. This would allow for expanding infrastructure size and operating volume to suit different greenhouse conditions, thus maximizing production. The robot is fitted with different sensors to enable autonomous navigation, perception, and to identify the environment and the operators (3D LiDAR, stereo cameras, and ultrasound). It also features the hardware necessary for cloud connection to share data in real time. All sensors have been validated to work correctly, hence the robot can move around the greenhouse. With the software currently used for collaborative robotics, the ultrasounds correctly identify the environment, and cameras and LiDAR can locate the farmer correctly. In this work, several gaps in greenhouse robotics are addressed by designing, developing, and validating a collaborative mobile robot with advanced sensors and algorithms with IoT integration. The robot lays the foundation for the implementation of autonomous navigation, collaborating with farmers in real-time and efficient operation in complex greenhouse environments, laying the groundwork for future advances in agricultural automation.

Keywords

• Mediterranean greenhouse
• CoBot
• 3D design
• ROS 2
• SLAM
• Sensors for navigation