Collaborative Workshops: AI Detection and Execution of Trachea Gripping Point

At the end of March and at the beginning of April there were held and series of workshops at NMBU in Ås (Norway). This workshop series consisted of a mixture of remote support and physical attendance and included attendance at NMBU by the following consortium members:

• ANIMALIA: Per Håkon Bjørnstad.
• OBUDAUNI: Kristóf Takács.
• BYTEMOTION: Haris Hadzic,
• NMBU: Alex Mason, Steven Ross, Luis Eduardo Cordova-Lopez, Ian de Medeiros Esper, Dmytro Romanov, Vebjørn Corneliussen Storvik.
• Remote support from CIKLUM: Anton Popov, Oleh Smolkin
• RobotNorge: Oddgeir Auklend

The list of testing activities shown below is a testament to the successful collaborative work carried out throughout this series. During this workshop there was a lot of testing and validation of physical system improvements. The highlight of the workshops was successful AI prediction and execution of trachea gripping point but picking of the saddle and placing onto the inspection rack is also a big deal.

Activities and achievements

• Implementation and testing of automated belly bar vacuum gripper.
• Tested automation of the CHU back vacuum.
• Implementation of Node-Red control for CHU vacuum grippers and spine skewer
• Tested CHU vacuum gripping during rotation of Axis 2 – without carcass escape.
• Implementation and test of new intestines pan to support hook pluck holder at the inspection rack.
• Addition of rapid tool changers on both robots for quick and safe shift between tools including the OBUDAUNI gripper, DMRI gripper, Smart sensorised knife and the rib cut saw.
• DMRI, OBUDAUNI and NMBU involved in fabrication of new tooling adaptors, designed by Oddgeir Auklend of RobotNorge to interface with the rapid tool changers.
• OBUDAUNI’s new upgraded gripper was successfully tested for functionality and payload capability on limbs.
• Imaging after rib cut carried out for AI modelling of organ connective tissue in readiness for automated cutting.
• AI generated trachea detection and gripping point prediction achieved.
• Trachea gripping point gripping execution achieved.
• Trachea cut path generated via manual robot control.
• Organ connective tissue cut along one side using manually controlled robot and replica sensorised knife.
• Trachea removal gripping path recorded
• Trachea, organs and intestines delivered to inspection rack by robot.
• Progress towards tool orientation and coordinate system alignment throughout the systems.
• Tested 3 different pick locations for the saddle whilst still on CHU.
• Saddle picked and placed on inspection rack saddle holder using DMRI gripper and manually driven robot.