Every year, hundreds of millions of parcels are managed in the Netherlands, many of which include returns. At the moment, this task still majorly relies on manual workers. If it is up to Fizyr, more and more robots will be enabled to handle parcels with automated pick and place solutions.

“It recognizes labels, even if they are at the bottom of the package. The optimal pick-up location is determined with a special depth image”

~ Sjoerd Bos, CTO Fizyr

Fizyr has developed a software solution that, thanks to ‘deep learning algorithms’, enables a robot to pick unknown objects. An envelope or a huge package can vary in size and shape, and yet Fizyr’s AI-driven vision software is able to determine what it is and where it should go, AD writes.

The software can then also locate the ideal picking point, allowing robots to pick the item without damaging or hitting other items in the vicinity. Fizyr can also determine the ideal pick-up sequence. “This makes picking and placing products faster, more efficient and easier,” explains the CEO Herbert ten Have.

Read the original article here