A Low Cost, Modular Robotics Tool Carrier For Precision Agriculture Research
Kjeld Jensen, Søren Hundevadt Nielsen, Anders Bøgild, Ole Jørgensen, Claes Lund Jæger-Hansen, Niels Jul Jacobsen, Rasmus Nyholm Jørgensen
- Year
- 2012
- Citations
- 30
Abstract
Current research within agricultural crop production focus on using<br/>autonomous robot technology to optimize the production efficiency, enhance<br/>sustainability and minimize tedious, monotonous and wearing tasks. But progress<br/>is slow partly because of the lack of flexible and low cost robotic platforms<br/>suitable for research within precision agriculture.<br/>This paper presents Armadillo, a $50k field robotic tool carrier with a modular<br/>design which makes the robot configurable and adaptable to a wide range of<br/>precision agriculture research projects. Armadillo weighs around 425 kg and<br/>consists of two 18x80 cm footprint track modules each with an integrated 3.5 kW<br/> electric motor, gear and motor controller. The track modules are mounted on<br/>the side of an exchangeable tool platform which allows an adjustable width and<br/>clearing height of the robot. The 48 V lithium power pack lasts 10 hours of<br/>operation.<br/>Armadillos industrial grade Linux based FroboBox computer runs the<br/>FroboMind architecture which is based on the Robot Operating System (ROS) by<br/>Willow Garage. FroboMind is a novel generic architecture that has been<br/>implemented and successfully tested on different field robots. It has been<br/>developed for research within precision agriculture and the design is highly<br/>modular in order to optimize with respect to extensibility, scalability, short<br/>development time and code reuse. All FroboMind software components used for<br/>the Armadillo robot are released as open-source.<br/>A prototype of Armadillo has been developed and tested as part of ongoing<br/>research projects within sustainable and organic weeding in row crops and<br/>orchards. Based on the experiences obtained, the first public version of Armadillo<br/>has been constructed and delivered to the University of Hohenheim, Germany.<br/>Current activities focus on further improving the Armadillo hardware and<br/>software and application to new research projects.<br/>
Keywords
Related papers
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Fractional Differential Equations
Igor Podlubný
2025
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991