If it can’t do everything, does having a robot add up?
An interview with Markus Frischeisen, Head of Automation Development at AUDI AG
Mr Frischeisen, as far as the car industry is concerned, where do you see the current challenges facing automation technology?
Well, one of the main challenges lies in the fact that structures are currently being dramatically altered. Up to now, manufacturing a car body has involved around 10,000 to 12,000 network connections. That’s standard.
But now the emphasis has switched to data. The buzzword is big data. We are adapting to it progressively. Basically, automated processes are becoming more data centric. These demand high-speed interfaces and, where devices are concerned, the ability to process larger volumes of data.
This is closely related to the issue of operability. Automation solutions need to be more intuitive if humans are going to understand and operate them quickly. It’s a bit like the swipe technology on smartphones. Ultimately though, big data mustn’t be allowed to become an end in itself. Preventive Maintenance represents a good case in point – logging huge amounts of data alone isn’t sufficient; the benefit for users needs to be the ability to predict the future. We want to know when the machine is going to fail if we’re going to be able to plan maintenance with any precision. To do this, FANUC has just introduced its Zero Downtime concept. We are really looking forward to implementing it and seeing how it moves us forward in terms of predictive maintenance.
One last point is the challenge posed by the persistent need to save energy. Where this is concerned we are always mindful to implement components that offer energy saving functions such as standby operation during breaks in production.
What kind of progress do you expect to see in terms of alternative processing and joining techniques, for aluminium and composites for instance?
Well, for a start all the standard joining techniques are used within the automobile industry. Bonding and riveting, for instance, are widely used solutions. These are easily achieved; especially with the robots and dedicated tooling that support this type of joining technology. The challenge is mostly posed by composites. Where these are concerned the future demands more precise and repeatable joining in order to increase quality standards.
Collaborative robots are currently a major talking point. How do you think this technology will change the automobile industry?
A well-known scientific belief is that 99% of activities can be automated. But as far as the automobile industry is concerned I think we need to be a little more differentiated. Firstly, the term collaborative robot invites a question – basically what do we understand ‘collaborative’ to mean? The word gives rise to different expectations in different people. Today, it is generally understood to mean passing a component to a human. But to me personally, as far as automation is concerned, collaboration means humans and robots working together in a car. We use collaborative robots in our production plant. They enable you to remove traditional safety fences and let humans and robots work together in the same space. But that’s only the first step. Right now I’m not aware of a standard solution for achieving the second step – having humans and robots work together on the same component. Plus, there is also the question of whether extensively automating processes with collaborative robots makes financial sense? Perhaps a human can do it better? Not forgetting the key issue of safety and workspace – it’s something we will always have to turn our minds to. In short the question remains: if it can’t do everything, does employing a robot to do the job add up?
Where do you see collaborative robots being used? And are today’s payloads and reaches sufficient?
If we approach the question from the perspective of ergonomics – i.e., collaborative robots doing the heavy lifting for us humans – at 35 kg, the FANUC CR-35iA is currently the only robot with a sufficient payload. None of the others come close. The technology needs to incorporate larger payloads and envelopes. Cost also plays a major role. How does it help me to have an extremely expensive robot that won’t ever pay for itself?
Intelligent, learning robots are also being discussed a lot. What role can they play in improving processes?
A big one. Right now collaborative and learning robots are hot topics within the world of traditional industrial robots and have a lot to offer. Nevertheless we want them to do more. For example, I can show a robot how to make something and have it incorporate this into its modular system. Perhaps, for instance, there’s a place in a box that the robot has problems reaching. I want to be able to show it how to pick the part and, when the next box arrives, have it pick it independently. What’s more, I want other robots doing the same job to access this information and to create a kind of hive intelligence without me having to do anything.
What kind of tasks can providers of automation solutions to the automobile industry expect to face over the next few years?
Off the top of my head, I’d say that automation requires people with the right skills. Here we’re talking about very different areas of specialisation – especially IT for big data, robot technology and artificial intelligence. Individual providers won’t be able to provide this kind of interdisciplinary approach. For that reason, partnerships will become increasingly important – with other manufacturers as well as with universities. Another thought that’s just come into my head is that Google, Apple and the like are currently buying up small robot manufacturers and have the financial wherewithal to buy in the skills they need. For the automation industry keeping up with them will be a mammoth task.
