Publicação
Design and control of cost-effective Agricultural Robotic Manipulators
| Resumo: | Global population growth and reduced availability of agricultural labour have increased the demand for research and innovative robotic solutions in agriculture. Cost-effective robotic manipulators are increasingly essential to address labour shortages and support sustainable agricultural practices. This thesis presents the design, study, and validation of a low-cost robotic manipulator configured for agricultural tasks. This work aims to create a system capable of performing targeted tasks efficiently and precisely comparable to human labour, enhancing performance and adaptability in variable conditions. The work involves selecting an adequate kinematic configuration, integrating advanced control systems—including a novel self-tuning feedforward component of a PI controller and reinforcement learning models—and incorporating sensory modules with stereo vision using monocular cameras for object detection and localization. A prototype SCARA manipulator with a 3 DoF PRR configuration was developed and tested to meet performance demands in unpredictable agricultural settings. The controllers and perception system were tested and validated. The results indicate that the researched methodologies suit low-cost agricultural manipulators. Furthermore, the manipulator prototype was tested in a laboratory environment, showing that the proposed manipulator is a viable low-cost solution capable of significantly reducing manual labour while supporting sustainable agricultural practices. |
|---|---|
| Autores principais: | Tinoco, Vítor Daniel Veloso |
| Assunto: | Manipulator Perception Control Agricultural Robotics |
| Ano: | 2025 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Trás-os-Montes e Alto Douro |
| Idioma: | inglês |
| Origem: | Repositório da UTAD |
| Resumo: | Global population growth and reduced availability of agricultural labour have increased the demand for research and innovative robotic solutions in agriculture. Cost-effective robotic manipulators are increasingly essential to address labour shortages and support sustainable agricultural practices. This thesis presents the design, study, and validation of a low-cost robotic manipulator configured for agricultural tasks. This work aims to create a system capable of performing targeted tasks efficiently and precisely comparable to human labour, enhancing performance and adaptability in variable conditions. The work involves selecting an adequate kinematic configuration, integrating advanced control systems—including a novel self-tuning feedforward component of a PI controller and reinforcement learning models—and incorporating sensory modules with stereo vision using monocular cameras for object detection and localization. A prototype SCARA manipulator with a 3 DoF PRR configuration was developed and tested to meet performance demands in unpredictable agricultural settings. The controllers and perception system were tested and validated. The results indicate that the researched methodologies suit low-cost agricultural manipulators. Furthermore, the manipulator prototype was tested in a laboratory environment, showing that the proposed manipulator is a viable low-cost solution capable of significantly reducing manual labour while supporting sustainable agricultural practices. |
|---|