A modular agricultural robotic system (MARS) for precision farming: Concept and implementation

Abstract Increasing global population, climate change, and shortage of labor pose significant challenges for meeting the food fiber demand, agricultural robots offer a promising solution to these challenges. This paper presents new robotic system architecture resulting modular (MARS) that is an autonomous, multi‐purpose, affordable platform in‐field plant high throughput phenotyping precision farming. There are five essential hardware modules (wheel module, connection robot controller, frame, power module) three optional (actuation sensing smart attachment). Various combinations can create different configurations specific tasks. The software was designed using Robot Operating System (ROS) with modules: control navigation vision module. A localization method dual Global Navigation Satellite antennas developed. Two line‐following algorithms were implemented as local planner ROS stack. Based on MARS design concept, two designs implemented: low‐cost, lightweight named mini heavy‐duty X. autonomous both X evaluated at traveling speeds payload levels, confirming satisfactory performances. further tested its performance accuracy in crop field, achieving over 537 m long path only 15% having error larger than 0.05 m. shown be useful field tests. makes easily adaptable tasks low‐cost feature it researchers growers.