Cofounder Johannes Trabert, who lives a poetic life in the Thuringian Forest, says the future will be focused on ‘the clever share of work’ between humans and robots.
ILMENAU, Germany—Johannes Trabert’s voice carries repose as he zips past the construction site for the latest Fraunhofer institute (plunked down alongside relics of East German industry here) in his Škoda and recites Wanderer’s Nightsong II, the poem Goethe scratched into a wooden wall of a lodge on top of the nearby Kickelhahn mountain in 1780:
Over all mountaintops
In all treetops
Barely a breath;
The little birds are silent in the woods.
Just wait, soon
You will rest too.
Trabert interprets the poem to mean “take a rest, cool down and rethink your priorities in life.”
That he did when a California company came calling for his expertise in robotics and automation. He chose not to wander from the satisfying life he has built here in the Thuringian Forest since he and all the Germans living behind the Iron Curtain were liberated in 1990.
That life includes founding MetraLabs with three friends from high school (while they were in college, in 2001), marrying a woman who also works in manufacturing, fathering twins, and helping companies like the German chipmaker Infineon vastly improve operations.
That life also includes recognition of MetraLabs’ contribution to the modernization of manufacturing—not only in Germany but also the Czech Republic, France, Great Britain, Singapore and the U.S.: The shiny new government research institute just a few blocks from his office will include a mobile robots innovation lab shared by Fraunhofer and MetraLabs.
Loads of time saved
Infineon is a prime example of MetraLabs’ influence.
The automotive supplier used to sometimes have to discard daily production because environmental parameters were not met—a discovery made after hours of gathering air-quality samples. That was using a human resource.
Since 2010, however, it has been using Scout, a MetraLabs autonomous robot that shrank to 10 from 300 the number of minutes it takes to collect the necessary data about the presence of rare gases in the air.
With 18 employees, MetraLabs makes robots that manufacturers commonly use for autonomous inventory taking that involves RFID transponders. “This is also something that is used in workshops, machine shops and little factories—where the robot is used to take inventory on expensive tools, for instance,” Trabert said.
Besides Scout, for clean-room applications, MetraLabs makes the robot Scitos X3, which it adapts for the goods a manufacturer wants to transport.
Infineon also uses a MetraLabs robot to transport wafers, masks and tools inside the factory. “In this case, the robot automatically puts the tools it delivers into the machine,” Trabert said.
The main goal is to take the human factor out of the clean room, he said, “because people are the source of contamination for all sorts of issues in the clean room.”
Collaboration with university lab, Fabmatics key
MetraLabs is succeeding in part because of its 15 years of joint academic research with the NIKR LAB- the Neuroinformatics and Cognitive Robotics Lab at Ilmenau University of Technology. The lab, headed by Horst-Michael-Groß, develops and advances theoretical methods for interaction between humans and mobile robots.
Also important is MetraLabs’ collaboration with Fabmatics. That Dresden, Germany-based firm specializes in the automation of material flows and handling processes in semiconductor manufacturing plants. “Our product fit well into their ‘catalogue’ of tools,” Trabert said, noting that the pair has implemented an installation at a chipmaker in the U.S. that he is not at liberty to name.
To be sure, MetraLabs’ products are not limited to clean-room applications.
The firm’s robots are also used at automotive suppliers in Germany. “This is a larger version of a Scitos X3,” he said. “This robot brings and takes larger and heavier interior equipment of a car from one machine to another. It’s like a fragmented conveyor belt, or an individual type of conveyor belt: The conveyor belt is substituted by several robots that carry a certain part, which is built up, step by step, from one work station to another work station.”
These robots see and interact with people.
“Sometimes they have robots with a touch-screen display, so you can gather information. Or they just wait until you go away. Or they drive around you. This is dependent on what the customer likes,” Trabert said.
Firm’s robots only need destination
MetraLabs’ robots are outfitted with an indoor localization system similar to GPS, making it possible for owners to set the machines to “make use of the autonomous capabilities and just drive around obstacles,” he said. “We actually don’t need a certain route; we just need a destination.”
The autonomous aspect of the robots includes 24/7 operation—without breaks for lunch, dinner or catching Zs.
MetraLabs robots capture data digitally and autonomously.
“The big question today is how to make use of Big Data—and how to get to the big pile of data that is necessary to derive certain patterns and to make use of the data,” Trabert said. “Therefore, we more and more develop robots that do a certain type of monitoring.”
MetraLabs first specialized in a robot that tooled around, checking the quality of the air. Later, the company developed robots for inventory based on RFID. Now, the firm makes robots outfitted with cameras and imaging processing—attributes that help manufacturers “keep track of what you are using and how much you are using and when you need to refill a certain box with a certain type of tool, for instance,” he said.
Customers have told Trabert that prior to using a robot, “they had cases where a really expensive tool just was hidden in a drawer in a table and a new one was bought,” Trabert said. “The tool was sitting right there all the time”—but humans working in the factory failed to detect it.
Classic AGVs are the competition
PAL Robotics is a competitor, Trabert acknowledged. “But for transportation tasks, to move goods inside a facility, our main competitors are the classic AGVs” (autonomous guided vehicles).
MetraLabs has a big advantage over classic AGVs because they “follow a certain track,” he said. “They have, for instance, a cable laid or integrated in the floor that they follow. Or they have a colored line drawn on the floor that they follow. Or they have some reflectors that they follow. That’s a very static track. There are no tolerances. Our robots, on the other hand, are tolerant against the environment, against the obstacles, against what is on the floor.
“You tell them their destination; they don’t need to have a certain route,” Trabert said. “Of course, you can define routes. You can define that the robot shall drive on the right-hand side if it’s a walkway. But you don’t need to change the infrastructure to use our robots. The robots learn the environment and create a map, using artificial intelligence (AI). They can adapt the map if the layout of the production facility is changed. And this adaptation is semiautomatic, so there is not much human interaction necessary to adapt for a changed layout and so on.”
Robots will handle monotonous work
The future will be focused on “the clever share of work” between humans and robots, he said.
That will leave the humans more satisfied by the work they do do, he added. “The places where our robots are installed create more attractive jobs or enhance the jobs that are there for humans—because the more human-related capabilities are used in a better way than before: The monotonous, stupid part of the work is then done by the robots.”
In places where human workers are not as plentiful as they used to be, “you need to make best use of them,” Trabert said. “And this is not carrying goods through the production facility; it is to involve them in production itself and the creation of value, checking quality and keeping everything running.”
Thuringia filled with prospects
MetraLabs sees a good part of its growth coming from new customers in the automotive supplier industry.
It cannot hurt that Thuringia is home to 640 automotive and supplier firms, said Arnulf Wulff, a senior VP at the State Development Corporation of Thuringia. “It is a chance for them to be very close to the market”—not only because Thuringia is thick with automotive suppliers but also because it is adjacent to the states of Hesse, Bavaria and Baden-Württemberg.
On top of that, the state’s thriving optics industry, which is intertwined with sensors and the digitization of manufacturing, provides a symbiosis that makes Thuringia “very strong and attractive,” he said.
AI wave is upon us
Automakers, which are themselves adapting to such market demands as autonomous vehicles for streets and highways, are adding AI capabilities to keep up with the competition.
“So a machine is knowing, ‘In five minutes I’ll run out of a part, and I need to tell the robot now because it needs four minutes to get me the new parts’,” Trabert said. “So there is more and more integration of intelligent use of data, and of course getting and using the data along the way the autonomous robots are driving.”
In just a few months, Trabert will be sharing insights like this from inside the advanced system technology branch of Fraunhofer IOSB and the Fraunhofer IIS wireless distribution systems department.
“That will be the glossy, glass-covered presentation hall for our robots,” he said, pointing to the halfway-done, four-story building. “It is a good contrast to our development area, which is a bit outdated. But it has charm, too, ja?”
* * *
Wandrer’s Nachtlied II
(Listen auf Deutsch: www.tinyurl.com/NightsongByGoethe)
Über allen Gipfeln
In allen Wipfeln
Kaum einen Hauch;
Die Vögelein schweigen im Walde.
Warte nur, balde
Ruhest du auch.