Automation and optimisation are becoming increasingly important in agricultural processes. In addition improvements in efficiency, soil protection is gaining attention. The project SOILAssist-2 addresses both topics. It aims at the development of practically usable methods and tools for optimal field management strategies with a focus on soil protection.
|Duration:||01.10.2018 till 30.09.2021|
|Donee:||German Research Center for Artificial Intelligence GmbH|
|Sponsor:||Federal Ministry of Education and Research|
Thünen-Institute (Inst. of Agricultural Technology, Staff unit for Soil Protection, Staff unit Climate Protection),
Christian-Albrecht University of Kiel (Institute of Physical Geography),
Osnabrück University (Institute of Computer Science)
|Application Field:||Agricultural Robotics|
Mobile autonomous, cooperative robots in complex value chains (08.2010- 11.2013)SmartBot
application-oriented solutions for agriculture, shipbuilding and small-scale industrial production (04.2011- 12.2014)SOILAssist
Sustainable protection and improvement of soil functions with intelligent land management strategies (07.2015- 06.2018)
In the context of the funding initiative “Soil as a Sustainable Resource for the Bioeconomy – BonaRes” of the BMBF, the DFKI Robotics Innovation Center developed an adaptive real-time assistance system for agriculture machines in the project SOILAssist. Driving and working strategies are planned on-line and provided to the driver by a driver assistance module.
The results of the first phase are now extended in the follow-up project SOILAssist-2: The main focus lies on the adaptation to further harvesting scenarios and the integration of a semantical environment representation.
The objective of the project is the spatial and temporal planning of the whole process and the coordination of several machines, taking a model of the soil qualities into account. The effects of the machines on the soil structure are of special importance. Theses impacts will be analyzed by the Christian-Albrecht University of Kiel. A first version of the soil model has been provided by Thünen Institute and will be extended in SOILAssist-2.
The environment representation will provide all information necessary for the planning, in particular semantically annotated 2D and 3D geometries. It will be the central source of data that can be inquired by all other modules and will ensure consistent and persisted data.
The project's results will be made publicly available in a modular, framework-independent open source library.