Adaptive Robot Gripper

Project number: 
Unilever UK/ACABI
Academic year: 
In the fast-moving consumer goods industry, manufacturers are motivated to increase the agility of their supply chains. This also means packing lines and packing machinery need to work with more complex product portfolios, without creating long periods of downtime between production runs. Adaptive robot grippers offer opportunities to minimize changeover time and increase agility of the equipment. These grippers can adapt to a range of product geometries and consistencies without mechanical modification.

This adaptive design is based on machine learning image recognition software, a sonar positioning system and a mechanical gripper. In addition, the team added emulation software based on RoboDK to deal with the technical challenges of sourcing a robot arm. A single Raspberry Pi computer runs the camera and gripper system. To keep the gripper from damaging objects, force feedback comes from a single sensor at the end of one side of the gripper. A sonar system determines the position of each object and feeds that into the Raspberry Pi to inform the gripper where to grab.

The mechanical gripper demonstrates the real-world use of the system, while the emulation software provides a comprehensive view into what implementation may look like on the factory floor.

Get started and sponsor a project now!

UA engineering students are ready to take your project from concept to reality.