Team Acorn
Team members
Simon Balzer, Justus Braun, Fabian Buitkamp, Jan Carstens, Philipp Gehricke, Eduard Gode, Frederik Hartmann, Isaak Ihorst, Jannik Jose, Andreas Klaas, Justus Klingner, Malte Klöpping, Julian Kronenberg, Luca Otto, Natalie Puls, Philipp Schein, Philipp Schlinge, Lena Schötker, Arthur Schreiber, Christopher Sieh, Marco Tassemeier, Nico Thiessen
Team captain’s name
Justus Braun, Fabian Buitkamp
Instructor(s)
Christopher Sieh
, Osnabrück University, Institute of Computer Science, Computer Engineering Group, Osnabrück, Germany
and Osnabotics e. V., Osnabrück, Germany
Eduard Gode, Osnabrück University of Applied Sciences, Autonome, kollaborative Agrar‒ und Sensorsysteme, Osnabrück, Germany and Osnabotics e. V., Osnabrück, Germany
Participating in the FRE since
Since 2004 yearly with varying team names
Description of the team and robot
Acorn is the name of our primary robot for the FRE and the team behind it. Started in 2004 at the University of Applied Sciences Osnabrück, the team expanded in 2019 when the University of Osnabrück joined. Last year, we founded the non‒profit Osnabotics e. V. to advance robotics in Osnabrück and the surrounding region through knowledge sharing and supporting student teams. We are excited for our first FRE as Osnabotics e. V. here at the DLG Feldtage, where our robots will tackle new tasks for the agricultural industry.
Robot specifications
W x L x H (cm):
45 x 90 x 64
Weight (kg):
22
Commercial or prototype:
Prototype
Total no. of wheels / no. driven wheels:
4 driven wheels
Drivetrain concept/ max. speed (m/s):
Differential drive, >1.4 m/s
Turning radius (cm):
0
Battery type / capacity (Ah):
Li-Ion / 432 Wh
Total motor power (W):
400
No. of sensors internal/
external:
Sensor type:
Internal: IMU, Wheel Encoder
External: Ouster Lidar, 2x Realsense cameras
Controller system software description:
The compute unit is running ROS2 and gathers information via Ethernet or USB sensors processed in nodes. The nodes provide ROS actions, which are triggered and managed by State Machines and Behavior Trees. This abstraction enables us to quickly adapt to problems coming up during the event.
Apart from ROS, Acorn uses CAN, CUDA, TensorRT, PointCloudLibrary, OpenCV, ONNX, YOLO for communication, model inference and processing.
Controller system hardware description:
The main compute unit is a NVIDIA Jetson Orin AGX connected to 2 Arduino Nanos für fast pattern controls or visual status and detection signals.
Short strategy description for navigation and applications:
Acorn detects rows and plants using a 3D laser scanner and two 3D depth cameras. The field is evaluated from point cloud data to generate a drive path, which is then followed by a customized navigation controller. The robot follows this path, while the 3D cameras are used for detection tasks in the rows.
These are the commercial team sponsors & partners:
AMAZONEN-Werke H. Dreyer SE & Co. KG, Allied Vision Technologies GmbH, ioTech GmbH and additional donators featured on our website.