067 - Precision Agriculture - High precision coordinate and alignments transfer

In future precision agriculture, it is expected that each plant (e.g. corn) will get coordinates with very high accuracy. The type of plants and their coordinates will be stored in a database with additional agricultural information. Therefore, already during the sowing process each finger of the seeder comb/seed drill has to be coordinated with high accuracy in the relevant navigation reference frame. On the tractor usually a dual frequency multi-system professional GNSS receiver with differential corrections is used (e.g. RTK, PPP, N-RTK, etc..). Some of the systems may have also a MEMS or microOptic based inertial measuring unit. In order to avoid the cost for an additional high end package for the each farming tool (e.g. seeder), it could be possible to determine coordinates on the seeder by a typical lever arm solution, e.g. determining the transfer attitude (transfer alignment) by an integration of visual and inertial sensors and gyro based system making use of the known linear dimensions of the farming tool (e.g. seeder). Because the drawbar has a certain tolerance in the trailer hitch, an additional distance sensor (laser, ultra-sound) must be implemented. 
The problem is not only applicable to the tractor and farming tool configuration but also to other land machines, and more generally, to machine control (e.g. civil works) or to constrain and steer large mechanical ensemble (e.g. machinery on civil works or even for satellites).

The objective of the activity is to develop a relative high precision PNT-Attitude determination and transfer alignment system to determine each point of a  mechanical ensemble. The study will define a generic model for the use of sensor fusion in the control of a mechanical ensemble, in particular machine control use cases, and then tailor it and demonstrate an extremely accurate solution for Precision Agriculture (seeder comb with accuracy of +/- 5 mm 95% TBC).

The tasks to be performed will include:
-    Assessment of state-of-the-art PNT technologies and products considered in machine control and precision agriculture (high accuracy positioning, attitude determination, non-GNSS sensors such as INS, radar, vision, etc.) and consolidation of the use cases;
-    Mathematical description and modelling of the mechanical ensembles to control, accounting for their kinematic motion;
-     Preliminary design and trade-off of candidate solutions (e.g. Kalman filter), considering the mathematical models, available and potential sensors (CCD ultra-sound, radars, etc.), targeted requirements (absolute and relative position, attitude, exposure to vibration, etc.). The trade-offs will also consider the challenges relating to the synchronisation of the various sensors and of the targeted application (e.g. machine control, sowing process, etc.).  Simulation of achievable performances tailored to relevant and representative use cases;
-    Tailoring to the solution to Precision Agriculture (sowing process and seed comb), with the development of a breadboard and its demonstration in field trials (involving field work);
-    Testing and performance evaluation.

The main output of the activity will consist of:
- the generic solution of the relative kinematic attitude and coordinates determination problem, and its tailoring to Precision Agriculture; 
- the development of a Precision Agriculture breadboard and its testing in the field.