TrimBot2020
The TrimBot2020.[1] is a research project funded by the EU Horizon 2020 research and innovation framework. The aim of TrimBot2020 is to investigate the underlying robotics and computer vision technologies to prototype the next generation of intelligent gardening consumer robots. The project focus is on the development of intelligent outdoor hedge, rose and bush trimming capabilities, allowing the robot to navigate over varying garden terrain, approaching hedges to restore them to their ideal tidy state, and approaching topiary-styled bushes to restore them to their ideal shape.
Design and technologies[edit]
The design of the TrimBot2020 robot required a combination of robotics and [[3D computer vision] research and innovation activities. Original developments in 3D sensing of semi-regular surfaces with physical texture (overgrown plant surfaces), coping with outdoor lighting variations, identifying different objects and types of surfaces, self-localising and navigating around obstacles, visual servoing to potentially moving target plants, leaves and branches are required to deliver all this on a small battery-powered consumer-grade vehicle.
TrimBot2020 is developed on top of the Indego platform, that is the Robotic lawn mower of Bosch, by adding a robotic arm and custom trimming tools. The robot is equipped with a camera rig of five stereo cameras for mapping and navigation of the garden and with three pairs of stereo cameras on top of the robotic arm for 3d reconstruction of bush surfaces and rose stems. The control of the robotic arm is performed by using computer vision feedback[2].
End effectors and arm control[edit]
The TrimBot2020 prototype platform is equipped with a 6 DOF robotic arm and custom designed end-effectors for omnidirectional trimming and rose cutting. When the robots reaches a location in the garden close to a topiary bush, hedge or rose bush, the cameras mounted on the arm are used in combination with computer vision algorithms to reconstruct the 3d shape of the target object. Subsequently, an algorithm based on an approximation of the solution to the traveling salesman problem is adopted to minimize the path to be followed by the robotic arm in order to trim the bush to the desired shape[3].
Visual servoing[edit]
An innovative development outcome of the TrimBot2020 project is the integrationg of visual information processing with robot control and the realization of a visual servoing module to approach bushes. Localization algorithms suffer from pose estimation errors, which are solved in TrimBot2020 by using a deep neural network (DeepTAM[4]) to localize the robot in proximity of bushes and support the approach to the target location[5]
Consortium[edit]
The TrimBot2020 consortium is composed of eight partner institutions spread around Europe. They are:
- University of Edinburgh, UK (coordinator)
- University of Amsterdam, The Netherlands
- University of Groningen, The Netherlands
- Wageningen University and Research (ex Wageningen University + DLO), The Netherlands
- University of Freiburg, Germany
- Bosch GmbH, Germany
- ETH Zurich, Switzerland
References[edit]
- ↑ Strisciuglio, N. (20–21 June 2018). TrimBot2020: an outdoor robot for automatic gardening. 50th International Symposium on Robotics. VDE. ISBN 978-3-8007-4699-6.CS1 maint: Date format (link)
- ↑ Kaljaca, D. (2019). Automated Boxwood Topiary Trimming with a Robotic Arm and Integrated Stereo Vision (PDF). IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE.
- ↑ "Demo of automated bush trimming".
- ↑ Zhou, H. (2018). DeepTAM: Deep Tracking and Mapping (PDF). European Conference on Computer Vision (ECCV), 2018. Springer.
- ↑ "Visual servoing with DeepTAM".
External links[edit]
- TrimBot2020 website
- TrimBot2020 on EU H2020 CORDIS portal
Media related to Agricultural robots at Wikimedia Commons
| Main articles |  | |
|---|---|---|
| Types | ||
| Classifications | ||
| Locomotion | ||
| Research | ||
| Related | ||
This article "TrimBot2020" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:TrimBot2020. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
