Most of the rotary tables produced by Goizper S. Coop., a cooperative located in the Basque Country (Spain), are used in welding applications in the automotive industry. This product requires an accurate assembly and specific manipulation. In addition, the variety of sizes that goes through the assembly shopfloor makes difficult the automatization of the product assembly.
However, thanks to the Sharework project, Goizper is to prove how collaborative robotics can improve their operators’ ergonomics, while increasing productivity by handing the most repetitive movements to Collaborative Robots (cobots) that can assist workers in a shared space.
Cobots are robots intended for direct human robot interaction within a shared space, or where humans and robots are in close proximity. Cobot applications contrast with traditional industrial robot applications in which robots are isolated from human contact.
LOOKING FOR A FLEXIBLE AUTOMATION IN THE ASSEMBLY SHOPFLOOR
The big variety of products that Goizper produces, combined with low production, requires a huge flexibility in the working area. During the last years, flexibility has been prioritised and, indirectly, this kept the assembly shopfloor in Goizper quite manual.
However, currently, collaborative robotics present the potential to keep the assembly area as flexible as before but aiding operator in specific tasks.
OPERATORS ERGONOMICS’ IMPROVEMENT THROUGH COLLABORATIVE ROBOTICS
In the assembly process, there are some repetitive operations that require a big force from the operator. These operations, performed by the operator, sometimes are not carried out in an optimal position. Besides, some operators complain about back discomfort due to these operations.
During the last years several improvements have been carried out to improve operators’ ergonomics, such as the installation of an hydraulic elevator table. Sharework project will bring ergonomics’ improvement further by automatising part of the assembly shopfloor.
Goizper, by participating in the Sharework project and as one of the industrial demonstrators of the project’s robotics software, chose to validate Sharework collaborative robotic software modules in two tasks of the rotary tables assembly process.
The first task, bolts tightening, consists of the application of adhesive on the bolts, which are inserted and screwed. During this specific task, the operator will work collaborating with the cobot in a same space. The robot will help the operator to determine their position and dimension and apply the same amount of torque to each bolt.
For autonomous screwing of rotary tables, the Sharework system is enabled to detect screws inserted in the corresponding holes. The screws are detected with a Random forest classifier that is trained on manually annotated data generated by intelligent segmentation of the rotary disk. The process from generating training data to using the detector is shown in the video.
The other task to be automatised consists of input axis turning, rotating the input shaft while observing the motion of the followers through two circular openings. In this task the robot will support the operator to rotate the shaft adjusting the force applied to rotate it.
At this moment, the goal is improving operators’ physical movements Goizper expects operators back and arm injuries to be reduced with the introduction of collaborative technology. Accordingly, operators’ happiness at their working place is also expected to increase, and this is one of the fundamental pillars of a cooperative.
Goizper S. Coop., one of the leading technology suppliers in power transmission components, is participating in the Sharework project since November 2018. Goizper manufactures and supplies customized power transmission components to meet market needs in sectors like metal forming, automotive, aeronautics, packaging, construction, marine and machine tools.
Mikel Anasagasti Alberdi
Electronics engineer at Goizper S. Coop., working in R+D+I department since 2014. High experience in software development. Working in H2020 projects together with IK4 research center.