Logo des Repositoriums

Auflistung nach Autor:in "Lee, Ting Sheng"

1 - 2 von 2
  • Konferenzbeitrag
    Predictive task scheduler and ERP system for automated vegetable cultivation in an outdoor environment
    (44. GIL - Jahrestagung, Biodiversität fördern durch digitale Landwirtschaft, 2024) Maike, Simon; Abbas, Farooq; Lee, Ting Sheng; Kühnast, Marvin; Weber, Bettina; Becker, Rolf; Franko, Josef
    Automated spot farming is a promising approach to overcome the ecological and economical challenges in modern agriculture. This requires sophisticated robotic controls and data management. The AgriPV-Bot, as a full farming system for mixed vegetable farming, achieves this by extending a classical ERP (enterprise resource planning) system towards monitoring single plant cultivation. The task scheduler analyzes this data, determines the resulting horticultural process for each specific plant individually and monitors the process execution that is performed by robotics. This paper introduces the features of the ERP system as well as the strategy for the predictive task scheduler.
  • Konferenzbeitrag
    Robotic process control for multi-vegetable micro spot-farming using digital twin simulation
    (44. GIL - Jahrestagung, Biodiversität fördern durch digitale Landwirtschaft, 2024) Weber, Bettina; Chande, Sahil; Maike, Simon; Verbunt, Maarten; Lee, Ting Sheng; Becker, Rolf; Franko, Josef
    Current robotic approaches in smart farming are often limited to a specific task such as weeding or harvesting. Contrary to this, the AgriPV-Bot aims at sustainable and efficient micro spot full vegetable farming by focusing on mixed vegetable cultivation through automated horticultural processes. Such a holistic approach requires sophisticated robotic process control. This paper presents the development of the underlying state machine built in ROS SMACH to handle a variety of tasks within the system. All processes and interactions of sensors and actuators are first simulated on the digital twin software Gazebo before being deployed in the real environment. This allows for rapid iterations of software and reduces dependencies on season and crop availability regarding physical field tests.