KUKA robots challenge automation – a new way of collecting oil and garbage bins

For us, it's easy to pick up something unsorted from the trash can – maybe a handkerchief, candy or screws of different sizes. But this is not designed for KUKA robots. "Picking up trash" is a top priority in the automation industry. This is also very interesting for the development of robust algorithms for the machine learning discipline – which is why it was one of the topics of the KUKA press conference in Munich in 2018.

They met in the halls of the Munich Exhibition Hall: industry experts and interested journalists. They are being transformed into industry and shared services through app stores, through big data and smart data. They talk about the global concept of AI use and the different regulations in various countries. They discussed the proper ordering of data, access to data, and the use of sensitive bots to pick up bins.

KUKA robots challenge automation – a new way of collecting oil and garbage bins

KUKA robots challenge automation – a new way of collecting oil and garbage bins

Christian Tarragona, senior vice president of research and development at KUKA, said in his opening remarks: "Cobots are definitely the focus of this year's automatica." Sitting next to him were David Fuller, CTO of KUKA Switzerland·· Dr. Christian Baur, CEO of Swisslog Logistics Automation. As a trio, they faced questions from reporters at a press conference in Kuka.

Data is the new oil

"Those who shut themselves out of the software today have already lost," said Christian Tarragona, who thought in particular of the new LBR iisy, a sensitive, lightweight robot that combines themes such as human-robot collaboration, sensitivity and ease of use. David Fuller adds: "We have to focus on the fact that robots are very simple, safe, connected and smart to use in combination with software." "This applies not only to the field of cobots, but also to all industrial robots."

Intelligence bridges the gap to artificial intelligence and analyzes the right data. "Data is the new oil. The internet is full of standardized image file formats such as: JPEG or. PNG, which is useful for learning image-based object recognition. However, there is no uniform file format for critical processes such as picking or spot welding. For AI applications, you need all this data. We specialize in collecting this type of process data and optimizing performance through the application of artificial intelligence.

KUKA robot systems must learn from the human hand

In warehouses piled high with everyday consumer goods, driverless transport platforms push entire shelves to the output station. The robot is waiting to take the goods out of the box and assemble them according to the order. "In the field of artificial intelligence, it is necessary to have things in parallel. In general, it is necessary to consider what data is needed and how it will be analyzed. Because if it's not the right data, then an infinite amount of data doesn't help," says Dr. Christian Baur. "As long as I have the right data, I can solve specific problems, like binary selection."

What makes picking up trash so complicated? It is the interaction of hardware with the right software. In addition to a responsive robot, a camera and a matching gripper are needed to hold handkerchiefs, candies, and screws. Our human hands are clearly the strength of this task. So this must be the goal of the whole system to learn – not just for KUKA robots. Only then can the system effectively play and learn together, and then at some point know that it has to handle a glass bottle more carefully than a plastic bottle.

KUKA robots challenge automation – a new way of collecting oil and garbage bins

Sections viewed

Contact

working hours