Today's smart industry mainly has low-volume, high-mix, high-complexity production that requires flexible and intelligent automation. This requires a robot that can work safely with people, is quick to set up and is completely integrated in the control of the production process. That is the connected collaborative robot, or cobot for short. Innovatiecluster Drachten (ICD) set up a CCR project for it in 2017. 

Various companies explored the applications of cobots and AIVs (autonomous intelligent vehicles that navigate between employees on the work floor and are therefore also a kind of cobot). The follow-up project, CCR 2.0, started at the beginning of this year. In it, six ICD members (BD, Neopost, Photonis, Resato, Variass and Ventura) work together with two northern universities of applied sciences (Hanze and NHL Stenden).

Safety first

The goal is twofold, says Joost Krebbekx, program manager at ICD. 'We want to 'cobotize' diverse processes, in which safe collaboration with people is paramount. You already see cobots being widely used for pick & place tasks, but we extend it to processes such as soldering, screwing, sealing and assembly. I think that we are leading the way in this as the Northern Netherlands. In addition, one of the companies is also looking at the application of cobots in its end products. Secondly, we want to connect the cobots to the control system of the factory.'

changeover

An enthusiastic CCR participant is Variass, a system supplier and EMS specialist (electronic manufacturing services) that develops and produces electronic and mechatronic solutions in Veendam and Drachten. The company already uses cobots for soldering in the assembly of electronics, says Jan Betten, head of Variass Development Support (VDE). 'In CCR 1.0 we learned how to make the cobot function as part of the factory and how to keep it safe. And we have learned to understand the business case. We are slowly moving towards the production of unique products, single-piece flow, and that means that we often have to make changes. This takes a lot of time with traditional industrial robots. We want to reduce the changeover time, have the changeover performed by lower-skilled staff and – the ultimate goal – have a robot change itself. For the assembly of electronics this means, for example, that the robot can teach itself to different printed circuit boards and the solder connections that it has to make on them.' The approach of CCR 2.0 for Variass is therefore that the robot must be connected to the factory. He must be able to read CAD/CAM drawings and be connected to the ERP and MES system. 'The robot receives its work instructions from ERP, and in MES we can read the status of orders that the robot executes and the robot's occupation.'

Valuation

In the CCNR project, Jan Betten particularly appreciates the collaboration with the ICD partners because they are working on completely different processes and sometimes already have a lot of experience with robots. Moreover, they look at certain things in a different way. 'As participating companies, we organize meetings where we exchange knowledge about complex subjects, such as grippers, vision, deep learning, et cetera. Together we will also have easier access to knowledge institutions and suppliers who would like to give presentations to several companies at the same time. The universities of applied sciences are good at robot and sensor technology and vision.'

The universities of applied sciences contribute their expertise, but in CCR it is mainly the companies that learn from each other, agrees Krebbekx. 'Early next year we will exchange experiences with companies on the Brainport Industries Campus in Eindhoven.'

This article was published in LINK Magazine of October 2019